3 Maximizing Average Prediction Change.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Prediction Progress #\n",
    "\n",
    "We now consider an agent whose objective is not to predict correctly the class of the stimulus but to maximize some form of prediction progress. In this case, although the agent keeps receiving feedback about whether or not it predicted correctly a given stimulus, the agent does not take that into account to decide what tasks to solve. Instead we are interested in an agent that tracks the evolution of its predictions in order to decide which task to perform. \n",
    "\n",
    "The agent now considers how good or bad it is at predicting the classes of a given stimulus favoring those tasks in which it is making some form of progress. For instance, if an agent performs very well in certain task but has reached certain level in which no matter how well it performs it cannot improve its predictions any further, then the agent should proceed with another task where it can notice an increase in performance. Similarly, if the agent performs poorly in a task and continues to do so at the same level, then it should switch to another task. Basically, in this case we want an agent that maximizes a metric of progress as opposed to an external reward given after classifying correctly an stimulus.\n",
    "\n",
    "More formally, in previous notebooks the estimates $\\hat{Q(a)}$ referred to the expected reward $\\mathbb{E}[R_a]$. In our exercise this was considered as the probability of succeding in a task and obtained simply by counting the number of successes divided by the total number of times the task has been selected. Let us refer to the probability of succeding in task $a$ as $P(s_a)$, thus for our new agent the reward is given by:\n",
    "\n",
    "$$\n",
    "R_a = \\Delta P(s_a) = P(s_{a,t}) - P(s_{a,t-1})\n",
    "$$\n",
    "\n",
    "The difference between our current probability of success and our previous one in task $a$. In this case $t$ may not necessarily refer to the actual previous time step, but to the previous time the task $a$ was selected. The exploration strategies will now be guided by a new estimate $\\hat{Q}(a)$:\n",
    "\n",
    "$$\n",
    "\\hat{Q}(a) = \\dfrac{1}{N} \\sum_{i}^{N} R_{a,i}\n",
    "$$\n",
    "\n",
    "Where $N$ may be the entire history of rewards, or just a time window of previous observations."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "from __future__ import division\n",
    "from IPython.core.pylabtools import figsize\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import scipy.stats as stats\n",
    "from scipy.stats import multivariate_normal\n",
    "from matplotlib.ticker import MaxNLocator\n",
    "from collections import deque"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# Parameters\n",
    "TASKS = 4 # Number of tasks\n",
    "SLIDING_WINDOW = 5 # Controls the number of past observations to use\n",
    "TYPE_AVERAGE = 'exponential' # For the prediction observations (e.g. 'total', 'movAverage' or 'exponential')\n",
    "ABSOLUTE_PROGRESS = False # If True, it treats a decrease in prediction capacity also as progress\n",
    "\n",
    "# For plotting the evolution of the accumulated rewards\n",
    "def plotRew(rew, std, steps):\n",
    "    ax = plt.figure(figsize=(12,9)).gca()\n",
    "    labels = ['Random','Greedy','$\\epsilon$-greedy','TS','UCB']\n",
    "    colors = ['blue','red','black','orange','brown']\n",
    "    for r in range(len(rew)):\n",
    "        ax.plot(rew[r], '-', markersize=8, color=colors[r], alpha=0.5, label=labels[r], linewidth=2.5)\n",
    "        plt.fill_between(range(len(rew[r])), rew[r]-std[r], rew[r]+std[r], facecolor=colors[r], alpha=0.15)\n",
    "    box = ax.get_position()\n",
    "    ax.set_position([box.x0, box.y0, box.width * 0.8, box.height])\n",
    "    ax.legend(loc='center left', bbox_to_anchor=(1, 0.5))\n",
    "    plt.ylabel('Cumulative reward')\n",
    "    plt.xlabel('Time steps')\n",
    "    plt.xlim(0, steps-1)\n",
    "    plt.show()\n",
    "    \n",
    "def plotSelection(cpTask, coord, plotName, absolute=ABSOLUTE_PROGRESS):\n",
    "    #print(coord[st])\n",
    "    #ax = plt.figure().gca()\n",
    "    axarr[coord].set_ylim([-.75, .9])\n",
    "    if not absolute:\n",
    "        axarr[coord].plot(np.array(cpTask[0])[1:,1],np.diff(np.array(cpTask[0])[:,0]), '-o', markersize=8, color='blue', alpha=0.5, label='Task 1', linewidth=2.5)\n",
    "        axarr[coord].plot(np.array(cpTask[1])[1:,1],np.diff(np.array(cpTask[1])[:,0]), '-o', markersize=8, color='red', alpha=0.5, label='Task 2', linewidth=2.5)\n",
    "        axarr[coord].plot(np.array(cpTask[2])[1:,1],np.diff(np.array(cpTask[2])[:,0]), '-o', markersize=8, color='black', alpha=0.5, label='Task 3', linewidth=2.5)\n",
    "        axarr[coord].plot(np.array(cpTask[3])[1:,1],np.diff(np.array(cpTask[3])[:,0]), '-o', markersize=8, color='orange', alpha=0.5, label='Task 4', linewidth=2.5)\n",
    "    else:\n",
    "        axarr[coord].plot(np.array(cpTask[0])[1:,1],np.abs(np.diff(np.array(cpTask[0])[:,0])), '-o', markersize=8, color='blue', alpha=0.5, label='Task 1', linewidth=2.5)\n",
    "        axarr[coord].plot(np.array(cpTask[1])[1:,1],np.abs(np.diff(np.array(cpTask[1])[:,0])), '-o', markersize=8, color='red', alpha=0.5, label='Task 2', linewidth=2.5)\n",
    "        axarr[coord].plot(np.array(cpTask[2])[1:,1],np.abs(np.diff(np.array(cpTask[2])[:,0])), '-o', markersize=8, color='black', alpha=0.5, label='Task 3', linewidth=2.5)\n",
    "        axarr[coord].plot(np.array(cpTask[3])[1:,1],np.abs(np.diff(np.array(cpTask[3])[:,0])), '-o', markersize=8, color='orange', alpha=0.5, label='Task 4', linewidth=2.5)\n",
    "    axarr[coord].set_ylabel('Prediction Progress')\n",
    "    axarr[coord].set_xlabel('Time steps')\n",
    "    axarr[coord].set_title(plotName)\n",
    "    #axarr[coord].set_title(plotNames[st])\n",
    "    axarr[coord].legend(loc=3, fancybox=True, shadow=True, ncol=4)\n",
    "\n",
    "def initPriors():\n",
    "    classes = 2\n",
    "    # Initalize observations for all p(y) with ones so initial p(y)'s = 0.5\n",
    "    priors = np.ones((TASKS,classes))\n",
    "    return priors\n",
    "\n",
    "def initParameters():\n",
    "    # Initialize parameters needed for likelihoods p(x|y) with mu=[0,..0] cov=identity matrix\n",
    "    classes = 2\n",
    "    mu = [[0]*classes for i in range(TASKS)]\n",
    "    cov = [[0]*classes for i in range(TASKS)]\n",
    "    for task in range(TASKS):\n",
    "        for cl in range(classes):\n",
    "            # Prior gaussian mu and covariance\n",
    "            mu[task][cl] = np.zeros((classes)) #np.array([0,0])\n",
    "            cov[task][cl] = np.eye(classes) #np.array([[1,0],[0,1]])\n",
    "            np.fill_diagonal(cov[task][cl], 100)\n",
    "\n",
    "    return [mu, cov]\n",
    "\n",
    "def random(counts):\n",
    "    # Get number of bandits\n",
    "    numBandits = counts.shape[0]\n",
    "    return np.random.randint(0,numBandits)\n",
    "\n",
    "def eGreedy(counts, epsilon=0.1):\n",
    "    # Get number of bandits\n",
    "    numBandits = counts.shape[0]\n",
    "    # Select random choice\n",
    "    if np.random.random() < epsilon:\n",
    "        action = np.random.randint(0, numBandits)\n",
    "    # Or act greedily\n",
    "    else:\n",
    "        #print('greedy bastard ', counts[:,3])\n",
    "        action = np.argmax(counts[:,3])\n",
    "    return action\n",
    "\n",
    "def UCB1(counts, step):\n",
    "    # Get number of bandits\n",
    "    numBandits = counts.shape[0]\n",
    "    # Tunable hyperparameter\n",
    "    c = 1\n",
    "    # Store upper confidence bounds\n",
    "    ucbs = []\n",
    "    for i in range(numBandits):\n",
    "        # Get number of times the bandit has been selected (at least once since they've been initialized)\n",
    "        T = np.sum(counts[i,[0,1]])\n",
    "        # Get mean(t-1) success for this arm (#success / #times arm has been played)\n",
    "        mu = counts[i,0] / T #equivalent: mu = counts[i,2]\n",
    "        # Calculate UCB\n",
    "        # UCB_i(t) = mu_i(t-1) + c * sqrt[2 log t / T_i(t-1)]\n",
    "        ucb = mu + (c * np.sqrt( (2*np.log(step+1)) / T ))\n",
    "        ucbs.append(ucb)\n",
    "    action = np.argmax(ucbs)\n",
    "    return action\n",
    "\n",
    "def thompsonSampling(counts):\n",
    "    # Get number of bandits\n",
    "    numBandits = counts.shape[0]\n",
    "    # The prior is given by a beta distribution, so define prior on alpha and beta (i.e. Beta(a,B))\n",
    "    # The bandits were initialized for now assuming the prior parameters are the same for all bandits\n",
    "    alpha = 1\n",
    "    beta = 1\n",
    "    samples = []\n",
    "    for i in range(numBandits):\n",
    "        # Since it is conjugate we can just get the posterior with the counts and plug it into the distribution\n",
    "        # Sample from the posterior p(R|History)\n",
    "        posterior = np.random.beta(counts[i][0] + alpha, counts[i][1] + beta)\n",
    "        #posterior = np.random.beta(counts[i][0], counts[i][1])\n",
    "        samples.append(posterior)\n",
    "    action = np.argmax(samples)\n",
    "    return action\n",
    "\n",
    "def selectAction(counts, strategy, step):\n",
    "    if strategy == 'random':\n",
    "        action = random(counts)\n",
    "    elif strategy == 'greedy':\n",
    "        action = eGreedy(counts, epsilon=0)\n",
    "    elif strategy == 'eGreedy':\n",
    "        action = eGreedy(counts)\n",
    "    elif strategy == 'ucb':\n",
    "        action = UCB1(counts,step)\n",
    "    elif strategy == 'thompson':\n",
    "        action = thompsonSampling(counts)\n",
    "    return action\n",
    "\n",
    "# Once the agent has selected a task, this function is called to sample a data point from one of the two distributions\n",
    "# that are available for that task\n",
    "def getSample(action):\n",
    "    mu1 = [[0,0], [0,0], [10,0], [10,0]]\n",
    "    mu2 = [[0,0], [0,0], [20,20], [10,5]]\n",
    "    cov1 = [ [[1,0],[0,1]], [[40,10],[10,3]], [[100,0],[0,1]], [[100,0],[0,1]] ]\n",
    "    cov2 = [ [[1,0],[0,1]], [[40,-10],[-10,3]], [[1,0],[0,1]], [[100,0],[0,1]] ]\n",
    "    if np.random.random() > 0.5:\n",
    "        mu = mu1[action]\n",
    "        cov = cov1[action]\n",
    "        category = 0\n",
    "    else:\n",
    "        mu = mu2[action]\n",
    "        cov = cov2[action]\n",
    "        category = 1\n",
    "    sample = np.random.multivariate_normal(mu, cov)\n",
    "    return sample, category\n",
    "\n",
    "'''\n",
    "This function simply tells us P(y=Class1|data point) and P(y=Class2|data point)\n",
    " action - the task or arm to select\n",
    " category - type of stimulus\n",
    " X - properties of the stimulus\n",
    " priors - prior observations [Task, Class] P(y) (in this case a 4 x 2 matrix)\n",
    " params - parameters for likelihoods P(x|y) corresponding to mu and covariance matrix ([mu, cov] each of size [Task, Class])\n",
    "'''\n",
    "def getPosterior(action, category, X, priors, params):\n",
    "\n",
    "    mu, cov = params\n",
    "    # get parameters to sample from the distributions [TASKS][CLASSES]\n",
    "    # From first distribution\n",
    "    mu1 = mu[action][0]\n",
    "    cov1 = cov[action][0]\n",
    "    # From second distribution\n",
    "    mu2 = mu[action][1]\n",
    "    cov2 = cov[action][1]\n",
    "\n",
    "    # Likelihoods\n",
    "    # p(x|mu1,cov1) or p(x=Data point|y=Class 1)\n",
    "    pxgy1 = multivariate_normal.pdf(X, mean=mu1, cov=cov1, allow_singular=True)\n",
    "    # p(x=Data point|y=Class 2)\n",
    "    pxgy2 = multivariate_normal.pdf(X, mean=mu2, cov=cov2, allow_singular=True)\n",
    "\n",
    "    # Priors\n",
    "    # First get how many times this task has been observed in total\n",
    "    numObs = priors[action][0] + priors[action][1]\n",
    "    # p(y=Class {1,2}) (Number of times an object of class 1 (or 2) has been observed)\n",
    "    # [TASKS][CLASSES]\n",
    "    py1 = priors[action][0]/numObs #0.5\n",
    "    py2 = priors[action][1]/numObs #0.5\n",
    "    # Obtain marginal p(x)\n",
    "    px = (pxgy1 * py1) + (pxgy2 * py2)\n",
    "    # Posterior p(y|x)\n",
    "    py1gx = (pxgy1 * py1) / px\n",
    "    py2gx = (pxgy2 * py2) / px\n",
    "\n",
    "    return [py1gx, py2gx]\n",
    "\n",
    "def updateObs(action, category, X, priors, params, observations):\n",
    "    # Update number of times a class of monster has been observed\n",
    "    # [TASKS][CLASSES]\n",
    "    priors[action][category] += 1\n",
    "\n",
    "    # Get total number of observations\n",
    "    numObs = priors[action][0] + priors[action][1]\n",
    "    # Retrieve specific mean and covariance matrix to update\n",
    "    mus, covs = params\n",
    "\n",
    "    # Append new observations\n",
    "    observations.append([action, category, X[0], X[1]])\n",
    "    \n",
    "    # Factor out\n",
    "    data = np.array(observations)\n",
    "    # Extract observations that correspond to a particular action and category\n",
    "    data = data[np.where((data[:,0]==action) * (data[:,1]==category))]\n",
    "    # Get only the datapoint X=(X[0],X[1])=(x,y)\n",
    "    data = data[:,2:4]\n",
    "    \n",
    "    #Extract window\n",
    "    winData = data[-SLIDING_WINDOW:,:]\n",
    "    \n",
    "    # Get new mean\n",
    "    #newMu = np.mean(data, axis=0)\n",
    "    newMu = np.mean(winData, axis=0)\n",
    "\n",
    "    # Update mean\n",
    "    mus[action][category] = newMu\n",
    "    # Get new covariance\n",
    "    #newCov = np.cov(data.T)\n",
    "    newCov = np.cov(winData.T)\n",
    "\n",
    "    # If it's the first observation we cannot really get the covariance yet and will just generate a matrix of NaNs\n",
    "    # So only update the covariance if that's not the case otherwise keep the initial assumptions\n",
    "    if data.shape[0] > 2:\n",
    "        if not np.isnan(newCov).any():\n",
    "            # Update covariance\n",
    "            covs[action][category] = newCov\n",
    "\n",
    "    # Pack parameters\n",
    "    params = [mus, covs]\n",
    "    return priors, params, observations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def observe(action, priors, params, observations):\n",
    "    # Sample from the distribution of the task that has been selected in order to obtain a set of\n",
    "    # properties X = [x_1, x_2] which correspond to sampling from the dimensions of the 2D gaussian and its label\n",
    "    # category = y denoting from which of the gaussians it was sampled from\n",
    "    X, category = getSample(action)\n",
    "    # Get posterior distribution with predictions\n",
    "    # This contains the probability the data point belongs to class 1 or class 2\n",
    "    pyx = getPosterior(action, category, X, priors, params)\n",
    "    # Sample from posterior to select one of the two options\n",
    "    prediction = np.random.choice([0,1], p=pyx)\n",
    "    # Verify if prediction is true or not\n",
    "    if prediction == category:\n",
    "        predCorrect = 1\n",
    "    else:\n",
    "        predCorrect = 0\n",
    "    # Update observations according to sampleprio\n",
    "    priors, params, observations = updateObs(action, category, X, priors, params, observations)\n",
    "    return predCorrect, priors, params, observations\n",
    "\n",
    "# Updates probability of success based on either all observations, a simple moving average or an exponential moving average\n",
    "def updatePSuccess(counts, arm, obsWindow):\n",
    "    if TYPE_AVERAGE == 'total':\n",
    "        # Takes total number of successes / total number of times the arm has been selected\n",
    "        counts[arm, 2] = counts[arm, 0] / np.sum(counts[arm, [0,1]])\n",
    "    elif TYPE_AVERAGE == 'movAverage':\n",
    "        # Takes successes in window / size of the window\n",
    "        counts[arm, 2] = np.sum(np.array(obsWindow[arm])[-SLIDING_WINDOW:]) / len(np.array(obsWindow[arm])[-SLIDING_WINDOW:])\n",
    "    else:\n",
    "        # Exponential moving average\n",
    "        counts[arm, 2] = pd.ewma(np.array(obsWindow[arm]), SLIDING_WINDOW)[-1]\n",
    "    return counts\n",
    "\n",
    "def simulate(numBandits, steps, strategy='greedy'):\n",
    "    # Sliding window\n",
    "    # Store observations about correct or wrong predictions\n",
    "    obsWindow = [deque([0,1]) for _ in range(TASKS)] # Initialized with 1 success (reward 1), 1 failure (reward 0)\n",
    "    #obsWindow = [deque([1,0]), deque([1,0]), deque([1,0]), deque([1,0])] \n",
    "    # Initialize prior probabilities\n",
    "    priors = initPriors()\n",
    "    # Get initial means and covariances\n",
    "    params = initParameters()\n",
    "    observations = []\n",
    "    # Save total reward and action selected\n",
    "    statistics = []\n",
    "    totalPredCorr = 0\n",
    "    totalReward = 0\n",
    "    # Store observations about [successes, failures, prob of success, last change in probability of success]\n",
    "    counts = np.ones((numBandits, 4)) #counts = np.zeros((numBandits, 3))\n",
    "    # Stores total reward per task\n",
    "    rewTask = np.zeros(numBandits)\n",
    "\n",
    "    if strategy != 'ucb':\n",
    "        # Accordingly, set initial probabilities of success with 0.5 for each arm except UCB since we'll set it below\n",
    "        # ASSUMPTION: We also initialize change in probability of success to 0.5\n",
    "        counts[:,2:] = 0.5\n",
    "\n",
    "    # UCB1 requires additional initialization\n",
    "    if strategy == 'ucb':\n",
    "        # It needs to play each of the arms once\n",
    "        for arm in range(numBandits):\n",
    "            # Execute action and observe reward\n",
    "            predCorrect, priors, params, observations = observe(arm, priors, params, observations)\n",
    "            totalPredCorr += predCorrect\n",
    "            statistics.append([totalReward, arm, totalPredCorr])\n",
    "            # Update successes and failures for the bandit [(succ, fail) so 1-reward]\n",
    "            counts[arm, 1-predCorrect] += 1\n",
    "            \n",
    "            # Update sequence of observations of rewards\n",
    "            #obsWindow[arm].append(reward)\n",
    "            obsWindow[arm].append(predCorrect)\n",
    "            \n",
    "            # Update probability of success\n",
    "            #counts[arm, 2] = counts[arm, 0] / np.sum(counts[arm, [0,1]])\n",
    "            counts = updatePSuccess(counts, arm, obsWindow)\n",
    "            # Initialize expected change in probability of success to initial probability of success\n",
    "            counts[arm, 3] = counts[arm, 2]\n",
    "\n",
    "    # For statistics\n",
    "    actions = [] # Stores task selected by the strategy at each time step\n",
    "    probs = np.zeros((numBandits, steps)) # Stores prob.s of success for every task at each time step\n",
    "    succFail = [] # Stores cumulatives successes and failures\n",
    "    countSF = 0\n",
    "\n",
    "    # Exploration space dynamics\n",
    "    acT = np.zeros(TASKS)\n",
    "    expSpace = []\n",
    "\n",
    "    # Capability to predict (learning progress)\n",
    "    lpT = np.zeros(TASKS)\n",
    "    leProg = []\n",
    "\n",
    "    # To keep track in the change of probabilities and also the steps in which they change\n",
    "    # It stores [probability of success, step, last difference] (One for task)\n",
    "    pt1, pt2, pt3, pt4 = [], [], [], []\n",
    "    # Another ASSUMPTION start all probs to 0, so first time each task is selected the reward is the probability of success\n",
    "    # One for task, it holds [updated probability of success, step in which it is updated]\n",
    "    cpTask = [[[0,0]] for x in range(numBandits)]\n",
    "\n",
    "    for step in range(steps):\n",
    "\n",
    "        # Apply policy to get an action\n",
    "        action = selectAction(counts, strategy, step)\n",
    "        actions.append(action+1)\n",
    "\n",
    "        # Execute action and observe outcome\n",
    "        predCorrect, priors, params, observations = observe(action, priors, params, observations)\n",
    "        totalPredCorr += predCorrect\n",
    "\n",
    "        # Update successes and failures for the bandit [(succ, fail) so 1-reward]\n",
    "        counts[action, 1-predCorrect] += 1\n",
    "        # Store succ/fail\n",
    "        if not predCorrect:\n",
    "            countSF -= 1\n",
    "        else:\n",
    "            countSF += 1\n",
    "        succFail.append(countSF)\n",
    "        \n",
    "        # Update sequence of observations on whether or not predictions were right\n",
    "        #obsWindow[arm].append(reward)\n",
    "        obsWindow[action].append(predCorrect)\n",
    "            \n",
    "        # Update probability of success\n",
    "        #counts[action, 2] = counts[action, 0] / np.sum(counts[action,[0,1]])\n",
    "        counts = updatePSuccess(counts, action, obsWindow)\n",
    "        # Store current p of success probs[action, step] update all of them for simplicity\n",
    "        probs[:, step] = counts[:, 2]\n",
    "        \n",
    "        # Save task-specific statistics\n",
    "        #cpTask[action].append([counts[action,2], step])\n",
    "        cpTask[action].append([counts[action,2], step+1])\n",
    "\n",
    "        # Get differences and assign reward accordingly\n",
    "        if not ABSOLUTE_PROGRESS:\n",
    "            change = np.diff(np.array(cpTask[action])[:,0])\n",
    "        else:\n",
    "            change = np.abs(np.diff(np.array(cpTask[action])[:,0]))\n",
    "        reward = change[-1]\n",
    "\n",
    "        # Update total reward\n",
    "        totalReward += reward\n",
    "        # And total reward per task\n",
    "        rewTask[action] += reward\n",
    "        statistics.append([totalReward, action, totalPredCorr])\n",
    "\n",
    "        # Update average change in probability of success\n",
    "        # First get number of times the task has been selected (at least once since we updated above the selection of the action)\n",
    "        if strategy == 'ucb':\n",
    "            N_t = np.sum(counts[action,[0,1]]) - 3\n",
    "        else:\n",
    "            N_t = np.sum(counts[action,[0,1]]) - 2\n",
    "        counts[action, 3] = rewTask[action] / N_t\n",
    "\n",
    "        # Store stats\n",
    "        # Time windows for action\n",
    "        acT[action]+=1\n",
    "        # Time windows for prediction ability\n",
    "        lpT[action]+=predCorrect\n",
    "        # Store every 10 steps\n",
    "        if (step+1) % 10 == 0:\n",
    "            expSpace.append(acT/10)\n",
    "            acT[:] = 0\n",
    "            lpT[:] = 0\n",
    "\n",
    "    #for i in range(numBandits):\n",
    "        #print(''.join( [\"Task \", str(i+1),\": \", str(np.shape(np.array(cpTask[i]))[0]), \" \", str(np.shape(np.array(cpTask[i]))[0]/steps*100), \"%\" ] )) \n",
    "\n",
    "    return statistics, actions, probs, succFail, counts, expSpace, np.array(cpTask)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random\n",
      "Task 1 28.0\n",
      "Task 2 22.0\n",
      "Task 3 24.0\n",
      "Task 4 30.0\n",
      "('Total reward ', 3.1651759830099682)\n",
      "===============================================\n",
      "Greedy\n",
      "Task 1 9.0\n",
      "Task 2 25.0\n",
      "Task 3 35.0\n",
      "Task 4 35.0\n",
      "('Total reward ', 2.8519315815176598)\n",
      "===============================================\n",
      "e-Greedy\n",
      "Task 1 18.0\n",
      "Task 2 22.0\n",
      "Task 3 32.0\n",
      "Task 4 32.0\n",
      "('Total reward ', 3.1283237506244399)\n",
      "===============================================\n",
      "UCB1\n",
      "Task 1 12.0\n",
      "Task 2 26.0\n",
      "Task 3 48.0\n",
      "Task 4 18.0\n",
      "('Total reward ', 2.9320822852369313)\n",
      "===============================================\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABC0AAALXCAYAAACpVbpOAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzs3XlwXOd97vnv2wsaewMgae4LTFESJVESBW2WIYuyDFqi\nZdlO7PjasWeSzJ0sNU4u62bGiWfGN3Jya6ayuIpzEyfXTjxJ5Njj3Y4XOQYii5ZEyZYEUpRISRSX\n5g5uaOxAN073eeePtxvd2Bs70Hg+VSic7nP69CEosV885/d7X2OtRURERERERERksQks9AWIiIiI\niIiIiIxFoYWIiIiIiIiILEoKLURERERERERkUVJoISIiIiIiIiKLkkILEREREREREVmUFFqIiIiI\niIiIyKKk0EJE5oQx5nFjzFcW+jpEREREJmOMOW2MeWihr0NERlNoIbLMZD6U+40xPcaYNmPMPxpj\nKubgrewcnFNERESKlDHmPxhjfmmM6TXGXDbG/MIY83vz9PYWjV1EFiWFFiLLjwUetdZWAbcDO4HP\nLOwliYiIyHJmjPlDYB/w58Bqa+1q4HeBdxpjSsY4Xr/HiCwT+p9dZBmz1l4GmnHhBcaYPzbGnDDG\ndBtjjhpjPpg91hjzG8aY54wxf2mMiRtjThljHs7bX2+M+Xnmtc3Ayvz3MsY8ljlnhzHmaWPMjXn7\nThtj/ldjzOHM3ZV/MMasNsb8JHO+FmNMzZz/QERERGTeGWOiwOeA37PWftda2wdgrX3FWvsJa+2g\nMeafjDF/Z4x50hjTC+wyxqwzxnzHGHMlMy75/bxzmrxxzTVjzDeMMbV5+z9pjDmT2fd/5D2/xhjT\nZ4ypy3vujsx7BOflByIiwyi0EFmeDIAxZgPwMHA88/wJoNFaW40bPPyLMWZ13uvuBt4EVgB/AXw5\nb9/XgJcy+/4M+B/JlFkaY67P7P8DXJjxJPBDY0wo81oL/ArwHuB64P2ZY/4YWIX7t+oPZuePLiIi\nIovMO4AI8K+THPcx4M+stZXAC8APgUPAOuAhYK8xZnfm2D8AHgPeBawFOoAvABhjbgL+Fvj1zGvr\ngA0A1tpLwH7g1/Le95PA/2etTc/kDyki06PQQmT5McD3jTHdwFngMvAnANbab2c+rLHWfhMXZtyT\n99oz1tovW2st8ASw1hjzNmPMJuBO4LPWWs9a+yxuIJH1UeBH1tqnMh/4fwWUAfflHfPX1tqr1tqL\nwLPAL6y1h621SeB7uDYWERERKT4rgWvWWj/7hDHm+Ux1Zr8x5n7cDY7vW2tfyBxyK7DSWvtfrbUp\na20M+AfgP2T2/y7wf1prL1prPdzNmA9nqiU+DPzQWvuctXYQ+Cww9N7APwOfyFxHMHNOTS4uskAU\nWogsPxb4QKaaYhewHVfNgDHmfzDGHMoMEjqAW3CVE1mXhk5ibX9msxJ3l6LDWjuQd+yZvO11uIAk\n+1oLnAPW5x1zOW97YMTjROZ9REREpPi0Ayvz56mw1t5nra3N7Ms+fz7vNZuBddkxS2bc8hngbXn7\nv5e373UgBazGVV4MnSszpmnPO/cPgJuMMVuAJqDLWvvybP1hRWRqFFqILGPW2meAfwL+KlMt8ffA\n/wLUZQYKR8i0kkyiDag1xpTnPbc5b/tC/mNjjAE2Zp4fTyHvKyIiIkvfC0AS+OAkx+Wv7nEWiFlr\na/O+qq21j+btf3jE/vJMRWcbbhwCQGb8MnSTxlqbAL6Jq7b4BK66VEQWiEILEdmHu4tQiyuNvAYE\njDG/iau0mJS19gzwMvA5Y0zYGNMIPJp3yLeA9xlj3m2MCQN/iKueeH72/hgiIiKyFFlrO3HtG39r\njPlVY0yVMSZgjLkdGG9Z9heBHmPMp40xZcaYoDHmFmPMnZn9/x34vzI3ZTDGrDLGPJbZ923gUWNM\ndmWSP2X070VPAL+JmxdDrSEiC0ihhcgyZ629hvtg/izwedzdjku4wOK5/EMZvX55/uOP4+a/iAP/\nBdcPmn2PY7g7FX8NXAXeB7zfWpua6NImeW8REREpEtbavwT+M/Bp3DjkEi54+DS5mxw273gfd4Pk\nduAUbnzxJaA6c8j/g2vzaM7M4/UCbkJxrLWv4ypLvwZcxI1dzo24nueBNNBqrR22T0Tml3Gt5SIi\nIiIiIpJljHkK+Kq19v9d6GsRWc4UWoiIiIiIiOQxxtwN/BTYYK3tW+jrEVnO1B4iIiIiIiKSYYz5\nZ6AZ+E8KLEQWniotRERERERERGRRCi30BcwGY4ySFxERkTFYa7V88DzReERERGRsMxmPFE17iLVW\nX/Pw9Sd/8icLfg3L5Us/a/28i/VLP+v5+5L5t9B/58vlS/+O6GddrF/6eetnXYxfM1U0oYWIiIiI\niIiIFBeFFiIiIiIiIiKyKCm0kCnZtWvXQl/CsqGf9fzSz3v+6GctIjOlf0fmj37W80s/7/mjn/XS\nURSrhxhjbDH8OURERGaTMQariTjnjcYjIiIio810PKJKCxERERERERFZlBRaiIiIiIiIiMiipNBC\nRERERERERBYlhRYiIiIiIiIisigptBARERERERGRRUmhhYiIiIiIiIgsSgotRERERERERGRRUmgh\nIiIiIiIiIouSQgsRERERERERWZQWNLQwxjxsjHnTGHPcGPNHY+yPGmN+aIx5xRhzxBjzGwtwmSIi\nIlLENB4RERFZvIy1dmHe2JggcAx4D3ABeAn4mLX2jbxj/negylr7GWPMyszxq621qRHnsgv15xAR\nEVmsjDFYa81CX8dipvGIiIjI3JrpeGQhKy3uBk5Ya09baz3g68AHRhzjA9WZ7WqgfeQAQURERGQG\nNB4RERFZxBYytFgPnMt7fD7zXL6/AW4yxlwEDgP/aZ6uTURERJYHjUdEREQWsdACvnch9ZMPAwet\ntQ8aY7YCLcaY26y1PSMPfPzxx4e2d+3axa5du2brOkVERJaE/fv3s3///oW+jKVG4xEREZFZNNvj\nkYWc0+Je4HFr7cOZx58BfGvtn+cd8yPg/7bWHsg8fgr4I2vtyyPOpR5SERGRETSnxeQ0HhEREZlb\nS3lOi5eBbcaYLcaYEuCjwA9GHHMWNzEWxpjVwA3AqXm9ShERESlmGo+IiIgsYgvWHmKtTRljPgX8\nFAgCX7bWvmGM+Z3M/i8Cfwb8kzHmVcAAn7bWxhfqmkVERKS4aDwiIiKyuC1Ye8hsUjmmiIjIaGoP\nmV8aj4iIiIy2lNtDRERERERERETGpdBCRERERERERBYlhRYiIiIiIiIisigptBARERERERGRRWnB\nVg+ZbZ7n0fp8C/3xGAHr4Zsw5XX1NNzXRDgcnvG5W1paiMVieJ5HOBxm48aNQB3nzg3ieQHCYZ/6\n+nKamhpm/H4iIiJSPDzPo7Wlhf5YjIDn4YfDlNfX09CUG6OMNdaor6+nacQxczXWERERWayKZvWQ\nH331L2jc3E60qnTo+a7eJM+drmP3R/ZO+8Pc8zz27dtHe3s7paXu3Ol0mueeexPYQWPjZwgG3bmT\nyS7q6p5j797dGjyIiMiC0+oh82us1UM8z6N53z4a29uJluaNUZJJnqurY/fevQCjxhoAyWSSuro6\n9maOaf7WvjkZ64iIiMwlrR6S0bglnvsQtz5Yn2hlhMYtcVqfb5n2eVtaWojH48MGEadOteF5W/G8\nAU6dyp07EokSjzfS0tI67fcTERGR4tHa0kJjPD4ssACIRiI0xuO0trSMOdYAiEQixONxWlpaaH2+\nZfhYJ3ueWRjriIiILGZF0x4SrYyATUHvaUhcgsqtULaWaGWE/pOxaZ83FosRiURIp9O88sorpNNp\nursDhEIb6OmBw4djtLfDPfeAMS64iMX6Z+8PJiIiIktWfyxGNBKBy5fh3DlIp+HOOyEYJBqJ0B+L\nEcMFFB0dHZw8eRLf97nlllsoLy8nEokQi8UoXQfRrRFI90PX6+4GTdV1UFI347GOiIjIYlY0oQUA\nJgjJK+CnINEGZWsBCFhv2qf0PPdaYww9PT0ADA6WEgqB70Mi4ZFIuDFIKJR9TdEUsIiIiMgMBDLj\nCDwPenvdtu9DMDi0PztK8X2f3swxqVRq6Bye5xHIdp1YC172PLljZjLWERERWcyK7LdrA6Vr3OZg\nN6T6APDN9Hs8s/2hgUCAQCD740oDLqQwmXMnEvmv8af9fiIiIlI8/Ow8E5mQAnB3OvL25481hp73\nc2OJcDicG8uY/KFb7piZjHVEREQWs6IJLbp6k24jU10BwEAbnT0Jyuvqp33e+vp6kkl37mBmwFFe\nbkilEkCC0lJ37swhJBKd1NeXT/v9REREpHiU19fTlUxCXiBBJpDoTCQor68fGmsE84KNdCbYSCQS\n1NfXU15Xnxnr5M1jlpn0c6ZjHRERkcWsaEKL507XuQ/zYBmU1ADQ1X6BA7FaGu5rmvZ5m5qaqKur\nGzaYWLmymnD4JKWl5dTWunMnEm71kBUrDtDU1DDzP5CIiIgseQ1NTTxXV0dXXnUF6TRdySQHVqyg\noalpaKyR3xLi+z7JZJIVK1bQ1NREw31NbqzTl9cGYn26epMcOLNiRmMdERGRxaxoljwdHBzMrV3e\nfxq/4yjlVbU0vO/ThFfeNqPzZ9dOf+KJJ+js7GTt2rV86EMfAur44hcHSacDbNvm88gj5TQ1NWjJ\nMRERWRS05On8GmvJU3DjiNYnnqD/O98h4Pv473wn5Q0NNDQ1DY0ZPM/ju9/9Lv/8z/+M7/vs3LmT\n+++/n6YRx7Q+92P6j36JAD5+9GbKNz1Ew31NGnuIiMiiNdPxSNFMxBkOh7n3gT3uge/BG5+HdAJ6\nXoMZhhbhcJg9e/Zw+fJlzpw5w5YtW3jssccAiMWgowN27IA9e2b6pxAREZFiEw6Hubepya0eAvDx\nj8PWraOO2bNnD2+88QYAjz76KHfeeefo8zzwPljxintizXvgbY1zfv0iIiILqWjaQ4YJhKHmVrfd\nexIGu2bltCUlJQAMDg4OPVdd7b53zc5biIiISDHKr4Twxl7pI79awhvnGEzehJ42PfYxIiIiRaQ4\nQwuAup3uu7XQ8cqsnDISiQAMTcwJEI267wotREREZFyzFloEciuI2NTYx4iIiBSR4g0tytbmVhLp\nODQ0w/ZMZCstxgotenqGJgMXERERGa6A0CIYDA4tezpuaAG5agtVWoiIyDJQvKEFQN0d7vtgJ/Se\nmvHpspUW+e0h2dDC911wISIiIjJKAaGFOyw36ea4FFqIiMgyUtyhRc0OCGTmGu04OOPT5YcWfqas\nIhtagFpEREREZByzGVpkxza+2kNERKT4FXdoESyF6E1uu+tNSPXP6HTZ9hDIVVvkhxbd3TM6vYiI\niBSrUN6Cbaq0EBERKVhxhxYAtZkWEZuGzldndKpspQWMHVqo0kJERETGZEyu2kKhhYiISMGKP7So\n2AyROrcdPzijCTnzQ4vsZJyRCJSWuucUWoiIiMi4phBapFITtH6YTNWGVg8REZFloPhDC2Ny1RaJ\nKzBwYdqnGqs9BLTsqYiIiBRAlRYiIiJTVvyhBUDtbbk1zeOHpn2asSotQKGFiIiIFEChhYiIyJQt\nj9AiXAVV29x212uQHpz4+HHkV1ootBAREZEpUWghIiIyZcsjtACoy7SIpAeh6+i0TjHWRJyQCy0G\nBmBwenmIiIiIFLvZCi205KmIiCwjyye0qNrmKi4AOg5O6xTjtYdUV+eOUbWFiIiIjEmVFiIiIlO2\nfEILE3BzWwD0nYPE1SmfYrKJOEGhhYiIiIxDoYWIiMiULWhoYYx52BjzpjHmuDHmj8Y5Zpcx5pAx\n5ogxZv+M3rB2Z267Y+oTcoZCIQIB9yMba04LUGghIiKy1MzbeGSKoYUdb5l2LXkqIiLLyIKFFsaY\nIPA3wMPATcDHjDHbRxxTA3wBeL+19hbgwzN608gKqNzitjsOgz+1OxTGmKFqi/zQoqrKrawKCi1E\nRESWknkdj0whtPB9H9/3x7loVVqIiMjysZCVFncDJ6y1p621HvB14AMjjvk48B1r7XkAa+21Gb9r\nbWZCzlQf9Lw15Zdn57XIbw8JBl1wAQotRERElpj5G49MIbRwh41znEILERFZRhYytFgPnMt7fD7z\nXL5tQJ0x5mljzMvGmE/O+F2j2yFY6rbjU5+QMxta5FdagJY9FRERWaLmbzwyW6GFVg8REZFlJLSA\n7z1Oo+YwYeAO4CGgHHjBGPMLa+3xkQc+/vjjQ9u7du1i165dY58xEIaaHdD+EvSegMEuKImOfewY\nsu0hgyPWNo1G4dw5hRYiIrJw9u/fz/79+xf6Mpaa+RuPZAOJdBp8HwKj7x2p0kJERJa62R6PLGRo\ncQHYmPd4I+7uRr5zwDVr7QAwYIx5BrgNmHCQMKm6O1xoYS10vAKrHyj4pZNVWnR3u9Nm57gQERGZ\nLyN/Sf7c5z63cBezdMzfeCQvkMDzIG8p9dwhCi1ERGRpm+3xyEK2h7wMbDPGbDHGlAAfBX4w4ph/\nBRqNMUFjTDlwD/D6jN+5bK37AreKyHizc49hrIk4IRdapNPQ1zfjKxQREZH5MX/jkZGhxZiHFBJa\nZFcP8d2XiIhIEVuwSgtrbcoY8yngp0AQ+LK19g1jzO9k9n/RWvumMebfgFcBH/h7a+3MQwtwy58O\ntMFgJ/TFoPLtBb1srIk4YfSyp5WVs3KVIiIiMofmdTwya6FFMO8PkAazoCvYi4iIzKmFbA/BWvsT\n4CcjnvviiMd/BfzVrL95zQ641OwmsYofnHJoMV6lBbjQYv3IKbxERERkUZq38chchRaExz5ORESk\nCCzfaD5UBtGb3HbXG5DqL+hl+RNx2ry2kpGhhYiIiMgwcxZaiIiIFK/lG1oA1N7hvts0dL5W0Euy\nlRa+75NK5ZYaKy2FTJ6h0EJERERGm63QIpBXKKtlT0VEpMgt79CiYjNE6tx2x8GCJuTMVlrA8BYR\nY6C62m0rtBAREZFRQnlhQ2rssCGUd4wqLURERJZ7aGGMm5ATYOAyDFyc9CWRvOXJxpuMU6GFiIiI\njKL2EBERkSlb3qEFQO3tLrwANyHnJPJDi/Em41RoISIiIqPM9pKnAFbtISIiUtwUWoSroOp6t911\nBNKDEx4+XnsI5EKLvr5xxyIiIiKyXBUQWgQCAYLBYOYQVVqIiIgotIBci0g6Cd0TL7teSHsIQHf3\nrF2diIiIFIMCQgt3WDhzyHgTcSq0EBGR5UOhBUDVNghXuu1JWkQKaQ8BtYiIiIjICLMVWhitHiIi\nIsuHQgtwdyxqb3fbfWchcW3cQ/PbQyaqtFBoISIiIsMEg7l5tGYUWqjSQkRElg+FFlnZFhGAjkPj\nHjZRpUV2yVNQaCEiIiIjGJOrtlBoISIiUhCFFlmRFVCx2W13vAL+2IOAiSbiDIWgMtNlojktRERE\nZBSFFiIiIlOi0CJf3R3ue6oPet4a8xBjzFBwMbI9BLTsqYiIiExgNkKLgJY8FRGR5UOhRb7oTRDM\ntH9MMCFntkVkZKUFKLQQERGRCajSQkREZEoUWuQLhKFmh9vuPQHe2D0ehVZaWDsnVykiIiJL1WyH\nFuO0s4qIiBQLhRYj1WZaRKx1c1uMoZBKC8+DgYE5uUIRERFZqmYltFB7iIiILB8KLUYqWwtla9x2\n/NCY5RLZSouxQgutICIiIiLjUnuIiIjIlCi0GMmYXLXFYAf0xUYdkq20mKg9BBRaiIiIyAgKLURE\nRKZEocVYanbkZuaOHxq1u5D2EFBoISIiIiNMMbSwY02QZUwuuFB7iIiIFDmFFmMJlUH1drfd/Qak\nhk9OMdFEnBUVEMrkHQotREREZJgphBYAqdQ4oUQgE1poIk4RESlyCi3GU5dpEfFT0PnqsF0TVVoY\nk5vXQqGFiIiIDDPF0GLSFhG1h4iISJFTaDGeii1QUuu2Ow4Om5AzW2mRSqVIp0cPFvKXPRUREREZ\nMmuhRaasU6GFiIgUOYUW4zEG6na67YHLMNA2tCtbaQETT8ap0EJERESGyQYSvg9j3Phwh0yl0kJz\nWoiISHFTaDGR2ttdeAGu2iIjP7SYaDLO3t5xxyMiIiKyHOUFEuNVW6g9REREJEehxUTC1VC1zW13\nvgZpV1WRbQ+BiUMLa6GnZ86vUkRERJYKhRYiIiJTotBiMrWZCTnTSeh+HSi8PQTUIiIiIiJ5phha\njL96SGZOC1/tISIiUtwUWkymahuEK912/BBQeKUFKLQQERGRPKq0EBERmRKFFpMJBKHmNrfddwYS\n1yattMgueQoKLURERCSPQgsREZEpWdDQwhjzsDHmTWPMcWPMH01w3F3GmJQx5lfm8/qG1O7MbXcc\nmnQizpISKC932wotREREFrd5HY/MWmiRXfJU7SEiIlLcFiy0MMYEgb8BHgZuAj5mjNk+znF/Dvwb\nYOb1IrNKV0LFZrfd8Qol4eDQrrFCC8hVWyi0EBERWbzmfTyiSgsREZEpWchKi7uBE9ba09ZaD/g6\n8IExjvt94NvA1fm8uFHqMtUWqT4ig2eHnh6rPQRy81ootBAREVnU5nc8otBCRERkShYytFgPnMt7\nfD7z3BBjzHrcwOHvMk/Z+bm0MVTfBEHXFhLsOkww6AYL41VaKLQQERFZEuZ3PFJAaBEKhfIOGSe0\n0OohIiKyTIQmP2TOFPKBvw/4Y2utNcYYJijHfPzxx4e2d+3axa5du2Z6fcMFS6BmB7S/DD3HqSo1\ndPZNXmmRTEIiAaWls3s5IiIiI+3fv5/9+/cv9GUsNfM7HikgtDDGEA6H8TxPlRYiIrLkzPZ4ZCFD\niwvAxrzHG3F3N/I1AF934wNWAo8YYzxr7Q9Gnix/kDBnau9woYW1rK9op7OvbtJKC3DVFgotRERk\nro38Jflzn/vcwl3M0jG/45ECQgt3mEILERFZmmZ7PLKQ7SEvA9uMMVuMMSXAR4FhH/7W2rdba+ut\ntfW4PtLfG2uAMG/K1kLZagDWRtoAW3BoISIiIovS/I5H8lo/Jgst3CEKLUREZHlbsNDCWpsCPgX8\nFHgd+Ia19g1jzO8YY35noa5rQsa4agugIjxITahz0vYQUGghIiKyWM37eCQYdF8ww9BCS56KiMjy\nsJDtIVhrfwL8ZMRzXxzn2N+cl4uaTM2tcKmFYDDI2tI2usaptKishEAAfF+hhYiIyGI27+ORcBjS\n6dmrtLDW3VgREREpQgvZHrI0hcqgejvBYJBVJddID/aOeVggANXVbluhhYiIiAzJzmsxK6GFZSEX\nVxMREZlrCi2mo24noWCIgPGp9s+Me1i2RaS7e56uS0RERBa/2QgtAnnFslr2VEREiphCi+moqMcP\nu0SizpzO3OUYLRtaqNJCREREhsxmpQVoMk4RESlqCi2mwxgGSm8EoIwubP/FMQ/Lr7Tw/fm6OBER\nEVnUFFqIiIgUbEEn4pxrnufR2tJCfyxGwPPww2HK6+tpaGoaGgxM+9yVN2Pt9zEGUldfJFzxoVHH\nZEML34fe3twcFyIiIlKcPM+jpaWVWKwfzwsQDvvU15fT1NSQG3tklz1Njd/WEcocM+nqIaAVRERE\npKgVbWjheR7N+/bR2N5OtLR06Pmu8+dpPnqU3Xv3zii4CJbWEvfqWFESx48fhk2PQmD4+fJDiq4u\nhRYiIiLFbt++ZtrbGyktza19fv58F0ePNrN372439lClhYiISMGKtj2ktaWFxng8F1hk+jOikQiN\n8TitLS0zOn8kEqEtuRaA9GA/dL0+6phobryieS1ERESWgXg8F1hkV0WPRKLE4420tLS6JxRaiIiI\nFKxoQ4v+WIxoJALxOBw8CM89N1SGGY1E6I/FZnT+kpIS2gdXMOiHSaVTED846hiFFiIiIstLJBIl\nmYQ334Rf/CK3glgkEiUW63cPphBapFIp/LEmxspfPUShhYiIFLGiDS0C2YGAtbmZMPPWHg1MMFAo\nRCQSwRLgUnIN6XQa+s5Asn3YMaWlEIm4bYUWIiIiy4O1cPmy+376dO55z8sMu6YQWoALLkbJr7TQ\nkqciIlLEija08LMf9vkTSfT0jN4/TZFMGnEpsZZ0KnOHo+PQqOO07KmIiMjyUloKa10HKfF4bgwQ\nDvvZDfe9wNBizBYRtYeIiMgyMWloYYz5S2NMtTEmbIx5yhhzzRjzyfm4uJkor6+nK5l0A4OyMvdk\nptKiM5GgvL5+RucvKSkBoN8vZyCwyj3Z8Qr4wwcOCi1ERERmbqmMR5JJ94G/eTMEMqOs06chkeik\nvr7cPZEfWlg75nkUWoiIiDiFVFrsttZ2A48Cp4GtwP82lxc1Gxqamniurs4FF9lqi+5uuhIJDqxY\nQUNT04zOn620AOgKXec2vF7oOT7sOIUWIiIis2JJjEfq6p4jmewiEslVW1y92oUxB2hqanBPZAMJ\nayE9duAweWihJU9FRGR5KGTJ0+wxjwLfttZ2GWPGvi2wiITDYXbv3UtrSwv9qRSBri78QIDynTvZ\n/Su/MqPlTiFXaQHQZddB8Cykk9BxEKI3Du3LhhYDAzA4CHkvExERkcItifHI3r27aWlpJRbrp6ws\nQHe3T3V1OW9/+25CoczYI38M4nkQGj0cU6WFiIiIU0ho8UNjzJtAAvg9Y8zbMtuLXjgc5t49e+D2\n2+FLX3JPbt8+fLAwg3MbY7DWkvQsRG+BeKurtPC6IeyqO0auILJq1YzfWkREZDlaEuORcDjMnj33\nDj3euROefx4uXIBTp2DrVkaHFtk21hHnyR2i0EJERJavSdtDrLV/DLwTuNNaOwj0AR+Y6wubVatX\n5+5inD8/K6c0xgy1iCSTSai7w+2wFjoODx2nZU9FRERmbqmOR975zlyV5dNPZ6awGBlajGHS0CJ/\nyVOtHiIiIkWskIk4PwJ41tqUMeazwL8A6+b8ymZTMJhrLL1wYdZOm20RSSaTULYOyla7HR2HhibW\nyg8t8lZcFRERkSlYquORigq45x63ff48nDjB7IQWqrQQEZFlopCJOP+LtbbbGNMIPAR8Gfjvc3tZ\nc2DDBve9rW3CJcamIltpMTg4CMZAbabaIhmHvtMAVFW5XaBKCxERkRlYsuOR++6D7PzdTz8NNqTQ\nQkREpFBLPdvxAAAgAElEQVSFhBbZT8JHgb+31v4ImPmkEPMtG1r4Ply6NCunHFZpAVCzIzeI6DgE\nuCKPqir3lEILERGRaVuy45GyMnjHO9z2xYtw8uxshBZaPURERJaHQkKLC8aYLwEfBX5sjCkt8HWL\nSza0gFmb12JYpQVAqByi29121+uQGgByK64qtBAREZm2JT0euffe3HybB14MZ7tIVWkhIiIyiUI+\n7H8N+Dfc+uidQC2LcF30SVVX50oeZjm0GKq0gNyEnH4KOl8DcvNaKLQQERGZtiU9HiktdW0iAJfj\nYa5ezeyYdmiRN4RTaCEiIkWskNVD+oCrQGPmqRRwYi4vak4Yk6u2mKXJOEe1hwBU1ENJjdvOtIjk\nhxZ20a0oLyIisvgVw3jknnugvBz8YJjTpzNjgnFCi2AwiMlMijV2aGFy1RYKLUREpIgVsnrI48Cn\ngc9knioBvjKH1zR31q933zs7oadnxqcb1R4CmQk5d7rtgTYYaBsKLdJp6Oub8duKiIgsO8UwHikp\ngcZGSAfC9PfDlSuMG1oYY4aqLcYMLSC37KmWPBURkSJWSHvIh3DroPcBWGsvAFVzeVFzJn9ei1mo\ntshvD7H5JRS1t+eWDIkfHLbsqVpEREREpqUoxiN33QVl1S6MOH0a/OT4K5pNGlqo0kJERJaBQkKL\npLXWzz4wxlTM4fXMrXXrcmHCLMxrkW0P8X2fdDpvwFAShcrr3Hbna0SrcoMNhRYiIiLTUhTjkXAY\n3rnLhREDA3DyTYUWIiIiEykktPiWMeaLQI0x5reBp4B/mNvLmiMlJbB6tduehdAiW2kBI+a1AKjL\ntIikE9QE3hh6WqGFiIjItBTNeKThTkO4zLV2vHbQIz1O5jB5aJFpD9GSpyIiUsRCE+00bgaobwA3\nAj3A9cBnrbUt83Btc8Jbs4bWZ5+l/+BBAvE4fiRCeX09DU1Nw2bqLkS20gJcaFFRkXfTp+oGCFVA\nqo/S/oOEw7fieQotREREpqrYxiOhEGy+LsyJ11IMdHscOgR33jn6OFVazA3P82hpaSUW68fzAoTD\nPvX15TQ1NUx5LCgiInNvwtAi40lr7S1A81xfzFzzPI/m/ftpPHOGaCgEFy9CZSVd58/TfPQou/fu\nndKHVX6lxbDJOAECQai9Da4+j+k7zeqads5fXaHQQkREZHqKZjwCsG5zmPPHBwikPZ55Bm6/3YUZ\n+SafiDMTWvgKLQrleR779jXT3t5IaWlu0rHz57s4erSZvXt3K7gQEVlkJmwPsW52yVZjzN3zdD1z\nqrWlhUZrXWAB0N0NQDQSoTEep7VlajdsJmwPgdwqIsDmarf8qUILERGRqSm28QhAIBJm82YI+h7d\n3dDaOvoYtYfMvpaWVuLx4YEFQCQSJR5vpKVljL8IERFZUIXMaXEv8IIx5pQx5rXM16uz8ebGmIeN\nMW8aY44bY/5ojP2/bow5bIx51RhzwBhz60zerz8WI1pTk7uVkQktwAUX/bHYlM6X3x4yqtICoHQV\nVGwEYF3ZKxj8/LcUERGRwhXNeASAcJjVq6G6zAUSzz47evVTtYfMvlisn0gkiu/DsWPw4ovQ2+v2\nRSJRYrH+hb1AEREZpZD2kPfOxRsbY4LA3wDvAS4ALxljfmCtfSPvsFPAu6y1XcaYh4Ev4QYt0xLw\nPLd6SGUldHZCX9/o/VMwaaUFQO0d0HeOikgvdeHjtPfeQCo1ugRUREREJlQ04xEAwmECAbhpm8cv\nUu4X55dfhne8I/8QF1qkUuNUUii0mDLPC5BKwdGj0NHhnjt3DrZvz+0XEZHFpZB/mbvH+LowC+99\nN3DCWnvaWusBX8etvz7EWvuCtTbbUPFLYMNM3tDP9ihmKyRGhBT+DCfiHFP0ZgiWUBqBtaUHAVRt\nISIiMnVFMx4B3NqnwIbVHitXuqeeew7yCzcLbw9RaFEoa30OH84FFgDXroGfWUw3HPbHfqGIiCyY\nQkKLg8A14Hjm6xpwxhhz0BjTMIP3Xg+cy3t8PvPceP4n4MkZvB/l9fV0JZNDAwXy7lx0JhKU19dP\n6XwTTsSZFSyB6A4iEVgRPk6J6dG8FiIiIlNXNOMRYGgsEkh77Nrlnurrc+0KuUMKbQ/RnBaF6OmB\nEyfK6ehwA7HSUvd8Ou1CjESik/r68gW8QhERGUshTQotwLettT8FMMbsBj4M/CPwd7g7FNNhCz3Q\nGPMg8FvAO8c75vHHHx/a3rVrF7uyI4A8DU1NNB896lYPARdaWEvX4CAHVqxgd1NT4VdPgZUWAHU7\niVxoxRifNZHDdHU1Tul9RERECrF//37279+/0JcxV4pmPALkbqB4HjffDM88A1euwIEDcNddEIlo\n9ZDZFI/DE09AXV0DZWXNVFY2ctNNUV54wVVZXLzYxb33HqCpafdCX6qIyJI32+ORQkKLd1hr/+fs\nA2ttszHm89ba3zbGlEz0wklcADbmPd6Iu7sxTGayq78HHrbWdozcn5U/SBhPOBxm9969tP63/0Z/\nZycB38evqaH8hhvY3dQ05SWuAoEA4XAYz/MmDi3K1hOpfhtwhbWlB+nqfCdgpvReIiIikxn5S/Ln\nPve5hbuY2Vc04xFgWGhhDDz4IHzjGzAwAL/4BTzwQC60SKfTpNNpgsHgiIvSnBaFuHwZvvIVN29I\nMBjmU5/aTTjcSizWz4YNAa5d86muLuf3f1/LnYqIzIbZHo8UElq0ZWbS/jruN+1fAy5nJq6aSePf\ny8A2Y8wW4CLwUeBj+QcYYzYB3wU+Ya09MYP3GhIOh7n3oYdcjSDAb/wG1NZO+3yRSATP88ZvDwEw\nhuCqOwiH/w2IM9h5Btgy7fdcKjzPo7Wlhf5YjIDn4YfDlNfX0zCNgEhERJa9ohqP5IcWADfeCGvX\nQlsbvPAC3H03wz4rU6nUGKGFljydzLlz8NWvQiLhHr/rXfDgg2GMcfOo3n8/fPe7bt+FC7B16wJd\nqIiIjKuQ0OLjwJ8A3888PoD7MA/iBgzTYq1NGWM+Bfw0c64vW2vfMMb8Tmb/F4H/AtQCf2eMAfCs\ntTNfoz3bxAjulsYMQotsi8iElRYANbdSEmnB89KEew5S7KGF53k079tHY3s70byfd9f58zQfPcru\nvXsVXIiIyFQU13hkRGiRrbb42tfcL9gvvADV1bnPSc/zhs2l5V6kSouRPM+j9fkW+uMx2q95vNQa\npt+vp3Z1E488Eua++4Yff/31EAy6eS1ef12hhYjIYjRpaGGtvQp8yhhTlXnck7d7RncbrLU/AX4y\n4rkv5m3/R+A/zuQ9xlRWltvORu/TlB1ATFhpARAqZ7DsRug9SmnydUjvgWDpxK9ZwlpbWmiMx4lG\nIm52q3AYKiuJRiI0xuO0trRw7549C32ZIiKyRBTdeCR/UnBrwRi2bYMNG+D8edci8u53Dw8tRlFo\nMYzneTR/ax+Nm9sZrC7l9Quw52boHTjPRXOUu+7aCwy/YVJaCm9/Oxw/Dm++Ce97HwS06qmIyKIy\n6T/LxpgdxphDwFHgqDGm1Rhzy9xf2hwaWWkxA9nQYtJKCyBVdQcAg8kUtuO1Gb3vYtcfi7nAIhaD\nw4fd4vNHjkB/P9FIhP5YbKEvUURElpCiG4/kVxuOqLYAt/TpkSOThBaBzL0nX+0hAK3Pt9C4JU5/\nbylHXx/KgrinIcJH3xmn9fmWMV93003ue18fnD07jxcsIiIFKSRL/hLwn621m6y1m4A/zDy3dOVX\nWswwtCi4PQQoqX07iXQNvg/e5YMzet/FLuB57lZR/qf/tWvw0kvw1lsEensX7uJERGQpKq7xyBih\nBbi7/ps2ue2jR8NkCzknrbSwBS+CUrT64zG6OiIce8tSFz7OmtLX2HFLilWrIFoZoT8+9g2TG27I\nVVe8/vo8XrCIiBSkkDktyq21T2cfWGv3G2Mq5vCa5t5CtIcA0RrDm8md1Jc/TbKrjZKBNihbO6P3\nX6z8q1fhRKZat6TEzRty+bIbVF28iN/bC08/Dffd59Z1WyCe59HS0kIsFsPzPMLhMPX19TRpslAR\nkcWmuMYj44QWxsC73w3/9E9gbZizZ+G66yYJLQCsP/zxMmMtHD/mEaiAmtBpakousPptECm7Crix\nVsCOvXRseTls2ODx1FOtHD7cz7FjAUpKfOrry2lqatB4QERkgRUSWsSMMZ8FvoKbrfvXgVNzelVz\nLRTKzbo0j5UW0ShcSt7OlrL9JBOWqvghWF+EocWpU5SfPUtXKuVaRHbsgKoq2LgRTp2i88oVyquq\n4Oc/d20ju3bBHXe4v5N55Hke+/bto729ndK8lqHz589z9OhR9mqyUBGRxaS4xiPjhBYAW7ZAfT0c\nPhzm4kX38Tl2aJE3jLNp3Dyiy4/vw5NPwlsnw9x2ewd1pWdYsxpKwoCfu6nkm7E/0z3P4/DhZs6c\naSQUinL2rBuznT/fxdGjzezdq6VQRUQWUiGhxW8Cf4pb6gvgWeC35uyK5oMxrtqit3fW5rQoqNIi\nCkk/Soe3lUTyBHS+CmubIFBEH4RtbfCNb9CwZQvNly/TuHUr0aoqt6+ykq4bbuDADTewe9MmV3nR\n1wc//rGbceyhh2D7drxUal6WSm1paSEej1NaWkoqlcIYQzAYJBKJEI/HaWlpYY8mCxURWSyKazwy\nQWgBbm6LI0fC+D6cOVNIpUUKKJn961zk0mn43vfctFnpwFoq+DZr10A4O8LNhBadPQnK6+rHPEdL\nSyuhkAssAK5edWO2SCRKPN5IS0sre/bcO+1rzF/RJGA9fBOmvK6ehvtU1SkiUogJQwtjTAj4rrX2\nwXm6nvmTDS1mqT3E8zx83ycwwZTTFRWumKAteQfJ5AlIJ6DrDai9dUbXsGjE4/Av/wLJJOFQiN1/\n/ue0trWNCh92NzURDoXcCONnP3Ori7S3wze/ibdmDc1nz9Lo+3O+VGosFiMcDnPmzBnOnDkDwKpV\nq1i3bh3V1dXENFmoiMiiUJTjkVDeECw1eiLNTZtg69YQL73k7ge0t08WWiy/FUQ8D775Tbfyh8Hn\nkds7ONkTYtPgANFw2PWM+IN09SY5cGYFuz/SNOZ5YrF+qqqiRKPQ1eWm4Nq61d3jikSixGL9M7jG\n3Iom0a1545re8zR/6yi7P6KqThGRyUwYWmTWLveNMTXW2s75uqh5kf2FeJbaQ8C1iJTlz5cxgjFQ\nXQ3XOm6gL1kO9EPHweIILXp7XWDR1+ceP/ww4Z07uXfnzvFfs2MHbN/uWkSeeQb6+2l99lkaz5wh\nunq1m42swrUrz8VSqVeuXKG1tZW+7DUDly9f5vLly1RUVHD99deTSCSGtY6IiMwWN6dOK7FYP54X\nIBxWD/14inI8MkmlBcCDD4b5+tfd794vveRx//0jDljGoUUiAV/7Wm6+74ZNz/Lu28+BvZfWmKU/\ndppA8hp+JEL59fez+yPjVzV4nrvhtHKlCy0SCbd6S3bKrez+6ciuaBKtLAU/CSYMJkC0MkLjFrei\nyb0PqKpTRGQihbSH9AGvGWNaMtsA1lr7B3N3WfMgGy7MUqUFuBaRiUILcOWGHR1BLvbfxq28AL2n\nIRmHSN2MrmNBJZPw1a+6SguA+++He+4p7LWhENx7L9x+Oxw4QH9rK9FQyFVetLe7cKmkBEpK3FKp\nzc2wdi1UVrp5MiorXbBhTMGXm0gkeOqpp3j55ZeHAouKigrKy8u5du0a1lr6+vo4duwYn//857nl\nlltoaGhg/fr1mCm8j4jIeNycOs20tzdSWhodel499BMqrvFIAaHF5s1hVqxwH4dvvOHR0eHmtR4S\nyBvGLaNlT7P3SS5dco9v3Xqah3fsJxQAIqu590O/Dee/56pZS1fB9ROHAuGwD+TuZ8Hw0CK7fzr6\n4zGiWyPQfx56TkBJDdTcOhRc9J9UVaeIyGQKCS2+Q65/1OImv1r662rNUqVFfmhR6GScALHOO4AX\n3IOOQ7DmoRldx4JJp+Eb33C1qwA7d7ppz6eqtBQeeojAiy/Cq6/mzpdIDAuWAu3t8K//Ovy1gYAL\nLrIhxsjvmW1bUcGbx4/z5JNP0tPTQ01NDT09PWzdupWNGzcSCAQYHBykra2Nc+fOUVtbi+d5HDp0\niEOHDrFmzRoaGhq49dZbh/29y9Kgu9qymLS0tBKPu8Ailcp1CsxWD32RKq7xSAGhhVvRyoUWqZTH\nz38OH/xg3gHLsNKisxOeeCJ3n2TH9j4+cMt3CKatC3E2fQSCJRCqdAekJl9ivb6+nPPnuygpyQWI\n2anKEolO6uvLp329AetBqgd6T2ZO3Al9MajcmtsvIiITGje0MO6W8geBtwGvWmt/Om9XNR+yFRGz\n2B5S6GScAFe6V+GXbSQwcA46XoHVD4KZfvnhgrAWvv99OJWZvP366+H9759S1cNIfnW1WzB9wwY3\nUWcy6b4GB2FwEH+sOUN8H3p63Nc4uhIJnjx+nGPd3UOVG+9et45zwSBJ3ydw7RqUlFASibBm1Sq2\nb9/O+9//fl599VWOHTuG7/tcunSJH//4xzQ3N7Njxw4aGhpYt26dqi+WAN3VXngKjYaLxfqJRKKc\nP+9Wh163zv0TCjPvoS82RTseKSC0CAaDVFcHWLXKJ532OHwYGhtdGwOw7EKLq1fhK1+B7m73eOft\nlvfv+C6Bvszn/7o9ULbabYcyq+GmBsBPQ2D8lVWamho4erSZixcbAfcZ4XmQTHaxYsUBmpp2T/ua\nfQKu4sPm5Wt95yBcA5EV465oIiIiORNVWvwtcBPwPPBnxph7rLV/Oj+XNQ+yoUUy6X7pnWACzYlM\nt9LCWugruYOqgXPg9UDPcai+YVrXsCCshZ/+FF57zT3esAE+8pFp/xyzyuvr6Tp/nmhFhZvTIk9n\nIkH5ffe59pPeXhdSjPe9vx+sxbeWly5c4KlYjMG0G9CVA7s3buS2lStJ1dXRcuoUsdOn8XyfcCBA\nfW0tTdXVhP/939leVUVPfT2HLl/m4LlzdA4O4pWUcPDZZzn4y1+yduNGGu68kx07dkxafeF53rys\niiKj5d/Vzqe72vNDodFo2R75Cxfc44sXXdn/qlXD9wtQrOORAkILd1iYLVuStLV5WOtWC//VX83s\nHLbkaXG3h1y86FpC+jN53n33QdNtz2EuZyoYanZAbd48WtnQAiDdD4Gqcc8dDofZu3c3Tz7ZyrFj\n/fh+gLIyn/vvL6epaWb/PpWbbrq6u4mWh6BsLSSv4A16tL7wHNeStVzjOn72vS8sidVENI4RkYUy\nUWjxLuBWa23aGFMOPIdbaqw45DcuJhJQPr3Sv5ETcU4mmvc7U9y/margTyA9CPFDSyu0eP55t0wp\nuFs+H//48AHYNDU0NdF89CiN8TjRvBCgK5nkwIoV7H7kEfc+w5p6x5BOc+nUKX74/e9zobTUTQM+\nOMht69ez+7rrqPA86O0l3NvLnm3bRr8+lRqaV6MK9z9DY00NJ+NxWi9c4Ni1a1igLRDgR9//Ps3l\n5ey47jruvPVW1m7aBFVVeJEILYcOEbt0iYS1nDhwgHeHQnzwxhsJB90dn7lYFUVGy97VBjdpW3c3\n1Ne7riLd1Z57Co1GC4d9BgaGF/sdPw41Ne6fuJn00Beh4hyPTCG0qKhIsmGDO+bIEZfdv+1tLJtK\ni9On3aSb2YLWhx6Cxp1nMbGn3RORFbD+0eGVntn2EHAtIuHxQwtwP+fHHruXI0fcX8c73gHvfe8M\nL7z3NA2bB2n+WRmN20NEV1+P11tF87P/zp31g9R5gzzScB/B4NVFv5qI53k079tHY3v7nK/uJiIy\n0kShxaC17hPQWttviq0GPn/CzBmEFiMn4pyI53m89FIrL77oUvxk0ucDdwZIdx8j2f8Sgbedxg9V\nLf60/fBhaGlx29XV8MlPTvvnN1I4HGb33r1jJvm7C0zyPc/j5z//Oc8//zy+78PKldTW1vLoo4+y\ndevW4Qdb627bTFS1kfkeSKXYtmIF21asoDuZ5FBbGwfb2uhKJBhMJGh98UVaX3yRdVVV3LZmDT+L\nxehKJCgNhTgfj7Oyp4eDwSCn33yTvbfcQriujmht7ayviiKjZe9a9/Xlupnicdi61aO/v5UjR95S\ny8IcyoZG1sLrr7tfPG66yU1yt1xDo/r6cl56qYtsKTq4n8uJE1BfP7Me+iJUdOORoTvWL75IwPfx\nOzoo7+4edcfa8zyOHz/OuXPnqKk5Q1fXWWpq6vn3f2/i4x8PL4vQ4tgx+Na33L0EY2DPHrhrZz8c\n/zZYP28eixHVjvmVFqk+CmGMC7M7O91H/4ykE3D+e4RDAXY/dD+t7dfTf7Kd42+2ce/qKF4gxdaN\ntQQT56CiftGvJtLa0uJuKI1YUW0uVncTERlpotDiRmPMa3mPt+Y9ttbapb1OZ/4/ujOY16LQ9pBs\nefSVK43097tB6sWLHt/81yf5rXsOc9O2lQSq3oKKTXOWts9KT/nx47mJMEtL4ROfGF4+MgvC4fC0\nP/hOnjzJj370Izo6OgAIBALcd999PPDAA2P/GbMjlIqK0fvyWetaiTIhRnVPDw/09nJ/VxcnTpzg\n5SNHOH7uHDaZ5GJPDz8/fZoLPT2sq6piXVUV3sAApdZCKkW8q4uWI0fYs2IFGEM0GqV/cBDuustV\nrSz98fiik71rffFi7rlUyuOpp5qJRBrZsmU1nZ31wPJuWZgr2dCoo8P1pAO89Rbccov7z305tkI0\nNTXw1a82k0o1Ul4epbLSBWkXLnSxefPMeuiLUFGNR4bdsU4kXItqPE7Xs88Ou2Ptxg37iMViDA4O\nUl7eQ2XlVc6cOc8//uNR7r9/LxtXLo3VQ6bbVnD4sBtyZLt4P/Qh2HGLhdPfAy8zscXa90LZmtEv\nnkZoAW7u7s7O3Aru03bx32CwC4Dwpke4d+dd7vnvfYHb3r4a4q2Q6oe+M64qJLJqUa8m0h+LuQrY\ngQE4c8alzps3QyDgVneLLc7rFpHiMFFosX3ermIh5FdazCC0KHQizmx5dEVFlFDI3THovtbCh98V\nIhTcQEfHNVaE26Bi45yk7bPSU37+PHzzm270EArBxz6WqU+dfyMDGN/vo739NQKBPoKZ1ov169fz\n2GOPsXr16pm/oTEupCktzTWdAwHg+sxXV1cXBw8e5NDLL9PR3AwVFVxMp7mYSNCXSNDl+5QZQxnw\nfE8PDVVV1IVChDs7CRw9Cl/4gqsN37bNfW3ZAiUl6iGdBfX15Zw508Xly+6//YoKuHixlVSqkVQq\nwrVrYfr7XcHQcm5ZmCvZ0CiTJQKu++ryZVizZnm2QgQCYW68cTfQSnl5P29/e4AXXvCpqipn5crd\npNPh2ei4KxZFNR4Zdsc6GHSfqen0qDvWLS0txONxIpEIg4OD+L7P1q1w5UqEgYE4f/3XLfzFn96Z\nO/EirbSYblvBL38JP/mJ2w6F4Nd+LTNZ7dXn3TxgADU3Q92do17rXjS90CJ7D2NGlRZdb7hJ1gGq\nrht2jQHruQqZ6M0uuLA+dB3FK99G6/F+3jp1hID18E14UVXeBgYHXfJ/8qRbOQ7w4nFay8ro7+3l\nVOYmnsYnIjIXxg0trLWn5/E65t/I9pBpCoVCBINB0un0hJUW+T31paWZD8NEjJJgKf12I4nEZTfD\ndc9xqNw662n7jHvKr11zDaWe536B//CHXcK+APIDmEikmsuXD3PixC9JJDopKzvLu961nfe+973c\neeedBGY4MehURKNRHnzwQR544AGuxuMcO3aMeGZNtsGeHjo9j87MsRcDAf4ulYKeHqo9j0slJfQc\nO0ZdWRl1x4+zoryc2spK2LyZ5tZWGoNBojU1Q+81nz2kxbDqQ1NTAy0tzSQSjYRCUbZsAc/rp7c3\nQjp9lrKyrbz6Ktxzj/vPe7m2LMyV7HKCHR3D//05cQLKyzu5++7l1wpx7hyk02G2bbuXX/1V2LED\nDh6EH/zAdaw1N8Njjy30VS4OxTYeGbpjDbnJqy9ehN5eoitW0P/KK/DII8RiMSKRyFAQ39HRwYkT\nr1FWthrfX8mRIzHaLt3D2uyJF2loMdW2AmvhmWfg6ex0FRE3bdbmzbhVNy495XaU1ML6CVYsC0Rc\n64ifKmjZ06zKzFQY06608Hrhwg/ddqgMNnxg2DUOrRYSqoCaW6DzCJ6Xorn5KRpvrmL1zeupv86N\nFhbNPBc9PfivvOIqLDI833dzkFlLdO1a6mtqqL96VXNcyJzRTbzlbaJKi+I2S+0h4KotBgYGJmkP\nyf3yvGqVCy2C1qOtDbwVa1hRmvmfrf8ieF1QffOsrt2dH5pcuACxmPsRVFe7X7aPHOnnkUfG+ezv\n6Rk+Zfejj8KNN87atU1VNoDx/RSHDz9BZ6cLd0KhUsrL38P27Zu5++67F+z6AoEAGzZsIBKJkEgk\nuHLlCmfTacKXLuGl06R8n0Aw6G7rl5dzJZUivmYNrRUV7vZzLDa0NFpHdzc39vbyTEkJdZWVrFi1\nirp166hdtWpeekiLZdWHcNjd1b5ypZX+fndX+9KlE9x11/X4/lYuXQqSSLj/1Kur3WuWY8vCXGlq\nauDgwWY6O11oVF3tJkNNJLpoazvAe96z/FohTmYWPDDGzRMMsHMnHD3q9h08CDffnNsnxSOQP+lm\nXR20tbnt7m7o7ibQ1galpXivvw6lpdRGo3R2dmKtpb29nVSqnYsXQ9TU1PP9H+zid3dZDGbRhhZD\nIU1Pj5tFNBJxK46tWjWqrSC7MFl2nu+KCteFunYt7sbOucw8FiaYmceidOw3Bfc/V6jCtWhMo9Ki\nv38ai8tZCxd+4No+ANa9b9QEoOV19XT1nidaGYGSOqi9jdYDT9O4LUXEdhA2qwALmMUxz8WRI/Dj\nH1OeStGVSrnV3a6/ntZXXqExlSIaCJC4cIFwXR2gOS5kbmgiWFm+ocUstYeAm9diYGBgwvaQ/PLn\nTZvcZ+m1U+5/rmvtYY7Ybazf1EfI7wavD+Iv4w/c4z4AZ2GOg+wvYNa6oDyVcsFJb6+7wROLBQgG\nYVa9FnEAACAASURBVOPG3NfatRD0Ei6w6MzUCOzaBQ0NC3r3/eTJHi5dOsyJE8/Q3p4iEoFVq6rY\ntm0Pq1Zt5/Lln83p+49l5M/jrbe6uXYtxg03bGLTpk2sX7+ek7/4BZsGBkj5PretXs1ta9ZwtquL\nF0Ih7rzvPrq6uhjYsMGVXXZ0QDxO5+XLXB4c5PLgoPvLunQJXnsNEwxSXVPDpWPHuOp5RGtrufjG\nG4Q7OqgJhzGlpaPSZ3eNLcRiMTzPIxwOU19fT9MECXV+hU5vb24KkKXWQnHlCrS1ubvajY3wnvdA\naanP1asbSCbdjxXcnALZ0GI5tizMlf+fvTePjuO+7nw/VdXVO1BoNDZiIxaCq0hKBGVBEiRLskjb\niuVlHCdOJp5MljNxdmZOJnnzznmTxTNvTs6bJEy8JfY4sZ1xYseSt1iWBFqyJIoSF4E7CJIg9h3o\nfa3u2t4fv8bKDVwkWbS/5/TpRqO6u7q6+1f3fu/3fq+qqrz//Xvp7+8lHs9x990y09M2uZyftra9\n9Per7Nz5du/lW4tLl8T1unVLPsaSBE8+CZ/7nDDl/N734Ld+S+R4P8WdA3v5ertxI9TWisUnGoVs\nFluWIZVCnZqCXI5mSaLC62XWcZjTdXBJBAIm6fQk3/ve19ginaClpZZweZyyyrfvfV0NiyTN0JDw\nhioUhCOv1wuNjcglaUOhYPCZTx1g6vgwqm3gCqjs+FgrVVV7wHHBxHcWPSJYtxf89dd/8ZsgLRaU\nFgs+3cHgtbdfgfhxSF0Utyu2CyXFKnQ+sIeeb/bR3RITxIWqkSOMx51gLCbT3lIsqW43gCS/fT4X\n+Tw884wgLYDOtjZ6FIXumhq0YJBcMIgWDqNHo4y5XLSnUjA4iNHURP/YGCe+8hX0ixdvuhpuGAa9\nrx0gFxv+sWyXud34qYrg2lhUbKmqMMcqK8NQVfrHxqC3l68NDNC8c+dPj9kdjJ9c0sLlYtFc4hba\nQ2DJjPNaSosFebTHoyFJgrhwCq0UzAlU2UE3Ahwb7GB76whBaYxEpojfHIOJbwumfrUr9g1iIQHL\n55dGhnm94raoJNik0yKOOHeu9BjJ5OHRf6HBnKW8HIKP7sb77ne/rdX38fFxXn75GSKRCqJRyOUk\nFOVeFOUhZmb6GB5+EY9HjId4q0iUKx2PcPhB+vv/J/PzZ+ju3oyiKLR3dTF24QLeYpF13d1EvF6q\nWlv502WLaz6fJxqNEovFiMViSF/+MvXT08SiUfKZzOJ31bEsktEos8kkRz/9aQYjEZoVBW8wiCRJ\nlHs8eFWVZ7/5Td7zK79CKBTiqaeeIp/P41826WViYoK+vj72XYWhHh7O4XZrjI2JeFOSRAuF1/vO\naqF4442l252d4nr5b9LvF8FpPC6sRHT9p9MbbjfGxwVp5HbDH/+xyFs+9znBxT37LLS1Qdm1JxLe\nMchklorrGzas/F9FBezZI3KFZFIMavrAB976ffwp3jz4W1tJTkwI9YEkiQ+9ogLa2kgkk/ibm6G6\nmtbBQSZSKTwuF5quowEbgIhtM+o1mDVDWDbMzOSBEX7Y98+46sbYuXMnW7duXWEU/nbCVtWlBXY5\ndB0uXcKen0fftoO/+sej7JjK0Ony4vfDjg4onJqgZ7yP936iC1fqgnictgXCa1RTKiXZxE0oLUD8\nVtdMWhRiMP28uK2WQ/2VVQaqqrL3Y/tEQj4oEvJL8Qo2Nu+mvT2F4hQhN4WRi9E75iOXLzAUKflF\nvFVJ+8CAYE3TafG334/6gQ+wt6NjMbEe8nppXbcOtbGR9kwGxXEwRkfpOXKE7mCQupoaWkoFrxut\nhhuGQc8399O9PorWvqyq/uPSLnOb8VMVwfWRGxxEi0RE5bVYxHAcelIpuj0eNK+X4YkJWuvrf3rM\n7mBcl7SQJKkb+BOgZdn2juM4bW/ifr018PnEgnwb2kPg2qTFnj2d9PX1EIt1L7ZpNLbt4cSlN3hg\n/RlCoc0UCjK9F9qoCnk5craPzS0yh5/7Hjnrh1xI17ChtRJVtm+KbV5I0BKJJZLhrrtEhS8aTdDe\n7qexUfRZp9MgOTYbzz6NFRllDJiv2sK5009QOSUxN9fLxEQ3VVXaCiHIm1l913WdF154gWPHjpHL\nxTHNCgyjhrq6J3G767hwoQfL6qayUqOurpX5+da3rIVhuRohlRKxWW2tSnf3f+XChe8QiRxh06ZG\nVFXl/vvvv6aywefz0djYSGNjIwB2Xx+PlcYt5AyDWDJJbHqa6OwssUSCtCQRjURojsXwyjLE4zg+\nH0m3m5iiMDt1iS//5Rkk2WEmkiBt+PGU1REMBvH5fHg8Hs6dOEHy3DkeveuuyxQaxaLMpUuipQhE\n5SmdXuqu+nFpobhWhQJUTp8W27W3Qygkbi//TYZCGrkcpc8vSXX1T6c33E44zlI7RGur8B70+0Uy\n/vWvi9zl+9+Hj3/8J2N4zsKxgMtJC4Ddu0WbyMiIINy2bRPH7Scdd0o80rlnj/ACiMWWvC2AZKHA\nofp69v7O74CqsucjH6Hvz/6M2NAQnlQKikVkSUKzLB6pCPJoy0YOj8XwJ9MU3BKSUmRkZITBwUH+\n/q//mnUeDxtqa6mtrcW7fj0xBPG/VqXd7YK/tZXkK68sDfft7BSL7cQEiXQaty/I03/+FX5mahy5\nrAGrvomtd/txu8GLh4dSo0y/dpqmDe3grrjMI+KaUEuMw014WsAN+Fo4tigyWaWqUOOHhJ/F1XZL\nVVe0e9iSSmPbvBiTmuzDyCfpOXKR7g4TrS5Ea3U9re3zb37SXigIQ53e3qX7Nm0SErBgEBVWtHy0\nLoyDymbhwgV6h4fp1nW0YpHYwtQ1jwe/y4X32DG++nu/R/umTddVEfS+dqCkRPGKlmkjDd5q/B4X\n3tQxvvrXv0f7hk13jPpiUUVQKMDYmDC5r6j4aasNiMrq2bPIL764gvjsjcXoTibRFAXKypBKn/+N\nftfudNxJCp61KC2+BOwDjgM/ng2TN4sF0uI2KS2u3R6ism/f3staKu699zFCgbsZODPOpXMG+YLM\nqYsSFZX/N8GKg1R6LtJ38TDv3nQCOV7Ott3vQlGUGz5xLSRo5893A2KCSSAAxWKSlpZD/O7v7kVV\nRXKRTDgk/vlZjEv9JIMw6VpP/9aP4kgy0Sj09ubI5TSGh4VYpblZXMSxuL3Vd8dx6O/v59lnnyVd\nYvyrqrwUCvdRX78XSVKwrMOYZjeyrBGN6rjdKoOD4PUKEuWb3+zlox/tWrPE+kZbXxb8QkwTTp8W\n4h3ThMZGla1bP0Z1dZjf/u3Hbur9L6/I+VUVf1UVjVVVsH07iUyGhvXryR47xv0jI8QMg6hhEDMM\n5goFzqpz/LudJnmvQk5RMOoga2R49myCaLYJXC7iExNopsnxwUHU4eEVCo1H/sOv8OqrI0xPz6Oq\nISRJLBfLOb4FBc/NtJ7cLlyvQlHzyD50XezD7mUm88t/k4cP54hGZWTZZuNGP7/8y2+PV8eddHJZ\njvn5pYLdco+GzZuFAeWZM3Dhgrje8Y4aXnlzWCAtSur4yyBJwoTz858X3sff+x785m/CsmFVP6m4\nI+IRVVXZu2/fFX/re5f91tVAgH3/43+ItXVoCCMWQ43FaHUc9mgajimTi3pp9pRhZ4uECjNEJnOc\nGh2lWZZxvF4GBgaYdLnom58nFgxy92OPLRp7Xk9pd7vQ+cgj9HzmM8L/oKpKSKrKykiGw7yUNymL\nbiAd+R6aJOMxpqnVp5GH6kSyLFuUhy8Rjcsgbby+j8VquJYpLdbYbructFjzBJH5Q8IkFKDqPii7\nMTOaJZ8LL1Tuovf1Q3RvtND8Mno2jmq5oRjD7yl/85L20VH4zneWEkOPB973Prj77isetxWKoUAA\n7rmHXDSKls2imyaqacKRIxh1dfRMT9NtGLRlMrSWJrpdqyKeiw2jtamQvgS5CQCMxCV6epN0b/XQ\nVpmhdUPpee4A9UXu9Gm0gQFheg+iJbvkzfYTO07WccRs9BdegLk57IXCcOnEmYvH0WRZbJdK4RQK\nGFVV9ExN0W1Za/6u3cm40xQ8ayEtEo7jPPum78nbgYUP8C1QWoAIVK6mQHjoPaKv/k//9DBFthBL\naXzxh+vZvf5/8fHtOlrAg21nmR84SF3rNjS/dkPmTKqq8vu/v5cLF3opFHJUVcnU1Cwk40sJmiRB\nxelXqJg/Bh1AbS3bf/EX2B51MT4ulBjHji1V101TtA3U1S0F1Ler+p5MJvnBD37AhQsXFu9rbW3l\n13/9k/zGb/SSTGaorNQwzRz19RpzczqyPIbH0874+MIjNMbHc1y8KDiqBRXu6oumia/DzbS+LLzf\nVEocDxA+IQ0N4njeyvG4ZkWutpa9n/wkhywLf3Mz/kiExkgEkkkOp2d4X6efco/EkCozZZukCyZF\n2eZnt+T5+qERomkHrVDA5XJhWxZOoUDMtpmemiJ7rp8XXzxN0vAQN57Bo21BkkIoSiWOE0aSKpFl\nFxs36uTzeT7zmc8QjUbxLlsUVwfEb1Z/6gpn+vFxoamXZTRZpvvSJb766l/Qqt2PJ6iyMeKCY+pi\ne5iqqjyxsYr3bFD5XM6FiYu2ChU1nxcf5kIb2VtQ/n+zTy5vZ3/wcmVB26qa+PvfL7xnF9pEWlvv\n7DYRx1nys2htvbrJX2UlvOc98NxzIod44QVxrH7CccfEI6qqrqlyKuKGK2yXzcLAAJoygJU/hGLB\n5rJ1mG6DPbbNUC7H+VgM0+fjlGUxmk5DPM6Pnn+ejm3bqKiooKysjFgsxoEDB678GrcJ6oUL7O3s\npHdoiFxrK3JFBbaq4tS2Mja1h1iFC8ankPN91Hqj4jcxMwONDVA3CoqOZHuh7nHwN9zYiy+0hzi2\nUDFcQ/2wgNXtIddFfgZmXxK3PVViP28Qq30ucoYbLdyAnp5jLG7T3uLDmD9JzxsRurf6aAtdPWm/\n4SKCacKLL8Lrry+agNPSAh/+sAiQrrbPq+MTSUIuK0O3bcayWdorKsC26T1zhu5UCq28nPiyOOZa\nKgLZiEG8V3i8ldB7Pkp3SxJNkonrBTBzGI6H/oExiPTytf0DNG/Y+c5SXqTT8NJLyC+9tLKAusoF\ndoV5708ChofFSW9iYvEuf329MIJtbgZZRp6YEIF9PI6ez6O63fSePSvUF34/cQDHwbDtn1jfixud\n3PTjjrWQFj+SJOn/A74FLGbljuMcf9P26q3CghnnbTDihGsrLdaCujrYsiVHPK4xOwuGITEw6id3\nz24CzjlcMlj5OCT6ANBcfnIjcdjZDIH1lzlUr05SUjkVr9xKZ+cePvABlfvuu8JO9PYuzRnTNPil\nX8JT5qVNW0o24nGb0dGlQRcgyOH6kifWrRoY2rbNsWPHeOGFFxaPqd/v573vfS87duzgzBmJXbv2\nMjTUS3NzjvPnh6isbKW1VUWW24nFFAoFsd6L5xOLfj4vLgu95Kshcl7R+lJWpuHzCX80l+vy1pfl\nxzY91kc63cfUXCu2vQdZVhdbDTTt1o7HWipytlpKwuvqxAXIjb6G1hIAJ01lMI7XsjCKRXAcHAey\nBYeQS6Ks4JA3DSTbYrNngjPJLLs9EnHdxbSp47irSGSPMaUOE5c70YsqclEhNmHgcs8wfKGN3/6N\nrzAxNYc/WI7P58fr9eJyuXC5XESjUb72ta/xnve8h8M9/8hjG1JUtgWQSiRALDnKFz/1NBuqduK2\nbYqyzGQxTmNLaM2tUCuc6Zdlx4Zt8/J8nEORc1Rpr1ITlnk+GWJPWxtqqdK4AA/wyIDgOwL9wNlV\nL7JAXqjq5bdv0329P/oR3bOzaD7fiorgWk4u11NovN39wUPCZgZNg3B45X6fPnaA+uAwpy8apB2V\nz/5VK//5v+7B7b4zg4np6aVBTFdqDVmO++4THkNjY3D0qGgTWVC1/YTizo1HbhSBANx9Nx0b7+bY\nP/TjyUWZy9SRLx7ngbIydpWSx3PZLH+VzeIqJT/xeJzh0olblmU0TSOfz7N161YaGxtxuZbCwtui\noHMcOHoUVVHo2rUL9u0DRSEahX/6p5LHtwRlG2tpXCcjpxJw8mTpsYPgEZ4IWWcdVN1E26m6TDZh\nZtZEWng8S5Zn120PsU0Y/5aY3CLJ0PQRkG987VrtczEU8dBaU4HqbaF9mwulMMmxM3N0t2XQlBzx\ntA7FjeCuWDFhpPOBPezfv/+yIsLo6ChPP/00O3fuxLbtpc/yrrtQv/994VYN4o0//vjS/O/r7fOq\n+KRf09gQDtPe1oaSSsHoKLnJSTRJgnQaR9dFv2mDIJ8uUxE4DkSPYEeOQKh08NUyCDSTK76KFiiZ\nyhsZjNnDgsTZ4karL2c4PkFre/2azmtvu8lnoQCvvSYuhiHMd0HIfBaYskJhMU+x7/DkehFTU4Ks\nWF7p8HrhwQfp3LWLns99bpEks2UZvF706mrGgPaqKi729Qn1ha7jzM5iHDpETzRKt8uF5vcv+l5E\nRkf5/NNPs2XnTlTbvmNUrauxGB/b9lL//4YN4PW+IxU8ayEtuhCzl3avuv/R2787bzEWSIu3wIhz\nrXAcmS1bxHljchKwDPKmxpS9m0r3AJorubSxmUPOjsDYU6UdqYRACwTWY7gb6PnuP65IUqamQWqf\n4If9fTQ07ANW/TDPnxeN5SAazj/xiSuWPNvb/UxNJWlu1picFGaec3OCtLhVA8OZmRn+7d/+jckF\nEwVg586dvPe97100kDx2DBRF5e67u/iDP4C//3uYn1/Z8O04Yr/Emm/z+OMiOFp+WVBFLEDXRatH\nLqeRLB3mVAq2bBG3F1pfVieATXaR0bFxRssm+PaxPlL2PmRZZXoaPJ4bOx5XPZFeYyFdIdEsQVbF\nYl5UEijhMF6gkEwiSyBZFh4N/IpNWdHBZ5qs96j4KPARn4VsSNi2xYwLXBVFlICH2fQ0fcUXKTpB\nbKlIuFpCC5UhK3O0rJvBqDJwkDBtCdOWMW0Jw5IxJYnIsSN8/cQX6W6JMdXvYgIJSVaxcBOdnGdz\n2uLw6/20afWM5wd55C6dghlk/Y4duFU32fRpvvGFV3nkQ79FMBjE6/Xi8XgWiY/FCsTChBvA8HjY\nPzzMeKJIxlLxmAUqVIeDoyn65ubY19V1GXERCkE0bvHC2BCvGXEc2UaVZVpDJaLjGqa9hmXROztE\nzokjSza2I+OXQnTWXk6QXA25o0fRcstaq2RZRM9utzi5pFKi6lWSVi9cDEm6rkJjRX9wfhqKcZBk\nyhzYbBziG//PQbY1rcdS3Hjqm9h8/4Ooqk8E35Jr5bXsAklddt+1359pCm8GEK0hC3Hw8t9R1wNe\n7g6JdSSdm+Arf9vHf/z920ukvO1BagnX87NYDkmCD31ItImYJnz3u/DJTwqu6ycUd248cpPw+6Gx\n2c34aIgJ7sa90QdtUxCL4R0bY5css7tYpKG8nNlslvyyOMW2beLxOPl8ni9/+cu4XC4aGhpoaWmh\nvr6eb33rWySTyWsq6K6LqamlSkFnJygKMzNiKNlCbvaud0GIVtIHJ9A0rUToZiEwCITJFxQybR+8\nOcWba5lswswC1dd9yMKUrGRyDUqL2RdBLyX8NQ/fuBJkGVb7XLS2zy/9024iZx1AC2ZE8cEqQuIU\nVN0PsntxwsiBAweIxWJ4vV4KhQKWZeF2uzlx4gTpdBpd1+no6ADHYeK11+ibn2ffffeJ81RDA3zk\nI1BVdWP7vIxM927cSNnBg6IFKRSCUAg5Hoe5OfRsFtXrFdXzhqXjtHgON9Iw8V1IX8If0EjmkmjV\nbaIoJ8nI/nXg86NnYqKqfiEuSByXDNkkkq5BdgzN5aa7IUHvS0/R9dhHQPas+O5cjcSPJEb5/P/7\nNFu27bxp/7jrwrJEcfDll1cwYv5t20gWCiLhXiDt8nnw+UjoOv473dRofl4UTBcmAYA40XV1wQMP\ngM+HCitIsoHGRsIjI4QbGgRJpijIqRTMz6OnUqheL72zs0J5IcvgciFVVGDE4xzp6+PnMxkMXaex\nowPDsnjl5Zf5/Be/yLbu7itO4HsnQjYMUSXp719hqrtQhX6nKXiuS1o4jvPIW7Afbw9uc3tIsVjE\ncZzFZOpmsFCVD4UEaVG0VIpFkL0qkeJW8i4PTaEwGAkoJrDlZZl3ISYuseP0nhigu3IGzamGvAZu\njUTcR5nfw/t3xBgfPEBj4xNLgfxgL/Kx17FN8Luq6PyjP0O9yolruYFhdbUgLhIJSKeT1NXdnIGh\nYRi89NJLvP7669j2wjEI8eSTT9K2TE8+Pc1i68euXSK2WT4FYgGSJPI9x0nw4IN+urtXvp7jiPPF\naiLj0qUlVUapTW7VfspLCWDABdkx2utMctHj5PwNfPxd83zrxFdRfHvJxwvU3H2cPY+9vzRX/tpt\nIjdbDb9SC4ntyOimSTpgEQ7VgKoxmZqlWMijumQiskq26JAtZNE8LlqbGzl//iKVKQdbskBx8LtM\nalx5pCLUe1RUGYKqOMb11Ra6aZLXdXL5PNmSk7Ni24IKk6TFIMGFi0xunnUuC8zS/bYg+TS/gRyQ\n6J9OMm2M8ODOIi7VhZlM0H88TaCiEgeJcM7if33qEtX1HdhIOI6M6vHi9foYOXyYIV3HMzWFN5fD\n4/FwqqKCcx4vKVeQcW8AY9MOane48agqsUKBAzt28MSjjwrDANMEw0AayvOZ//aP5FxuNpU3Eq6w\nwbaZKBTom5lh3wc/iOo4GLrOgRMnGJ6awjAMZNumaI3xq90+Kn0uwWjbBpHkCJ8/fI4t4VpUF9cl\nMmR7lSLHtpfkQSBOxj09lz2ud3SU7slJtEBA9GiVLpqqCoXGN75BTp5G2+oCuwDpC+CA7djEJydY\nb5iM6h7usVNggz54hOGpZ2jv6kK5DiEBiO/1IpGx7LpEasQiKh0eFdvjYme1ClPif71H3qC7pg9N\n8UBepqNJQU/KeGWFQDHFq8/+K4++932l51kgSRQM07xh34+3NUhdhYXWkKoqoTy5HsJheOwx8dFH\noyKu2/sT6hF7R8cjt4Dm9QpTEyBLFueHVZyNfqRGv1g7JidRCwV8ikJLRQWO30/9zp0kEgkSiQTJ\n5FIhxDRNRkdHGR0dZWBggLGxMUKhEBUVFYvtJB6P57KWkmsqMo4dE08uy9DZydgY/PM/L/G/7363\nmKRumsvOYxUBqB2BAuimxWvygzy89yZH6CirSYu1IRgUpMU1lRaZEYi8Lm77G6D6oZvaxSthyeOi\nVIyQVWRPJfg86JkoqlsRQYqZXezNlR2D4eFhPB4P8XicM2fOYNs2qVSKQqGw2A5ELofV18fY2Bi9\nuRwD0Sg7H3qIpnvugddeuyWz1ivGI4EAejjMmM9HezAoPvxlakJbVSF5Hia/B6Yg7jt37aLnjfV0\nVztopdjJRkY3FcYKzbRv2cHFlw+iBXNC5YKDY2QhI2R9GpC7MAq1F8S5Qy0DVxmoZfT2nqe75iKa\nGoCiDooXw1Y5cvQEP78ljeHRaVzfAVy59eamyG/HEYnjD38oxhsvoL4e9uyhs7FRFB+mppYMa3Vd\ntAKHw+zds2dNx/8dh0QCXnoJTp1aak1SFEFwPvzwZaN7lpNkD5VaaltisUWfHtvrRdc0xmpqaG9t\n5eLRo4KwADBNnGSS3hdfpDudRgsEGB4dxWhooOfECbrzeR4EIseP09jR8Y71fViE42DPzAiSzFpm\nAbWsOPZOU/CsZXpIBcKt++HSXS8Bf+44TvKqD3qnYEFpUSyKD3SNFdHVWFBaOI6DYRiLJMbNYCEB\nDwbFspUotBJPTbDO68E0dSoqPMI9211Bwtbxb30Q2ndCdhSyI5AbB6tILh1Hq7dEr2V+BgcI5F3U\nusvwhssZnnwdI9dFz3e+RHf1OFrhHGw1QZKINHr4/Ff/+1UD+eUGhr29SwaGzc1+fuu3btzA8NKl\nSzzzzDPES+ZPsizzwAMP8O53v/uy51qIfyRpyVTxSpNZAAqFJOHwlUkUSRJrYTC40ghvdNRmfl4k\nFhMTl51bUVVbGES1eyDVD/lZZGDjOhu1cArTtJhPn6Gm8hW8XpX7dq5DvVhijmVVjK6VPUvXi7fd\n9B47SXdNP5riA10B2Q3qSunnlfxLriTRvJhtYV14kk11MrIkg6+ObffezdDgEGNT88TdLURshfp8\nlPc2byI5pXD4EASNPC5Tx6ckKVZ6kSorQLJAdhg33Wi+e5Bki3UBi3KvRblkk42fJl3Moyg2suyg\nyrZ4DDaSbBFEoUwxqZSK2I6N41g4OKSKRfy2je1AGSaGq0hjWMV2BBFXoRTweUuRrRdGp6Jsrpxb\n8d4dR6JlcwxvLIlSnsKxwVbdhIxxdlbYZJscOhQ/WuU8rriE5Ui4ZYX+Q4cJFF/D5w8uXi6MDtHY\nMIwkBwmFJRqbZWxHwSGIXjR4nXHuf7Cbr3z1/xBTk7g7QthITE6M8mhLgUteN52du1BcHgxT4siP\nDvPz/97CcFfQ2NSGZRic7LvIn7z+BuG6TrxuN03V1VT6bIqZCS6tK2Kni/jNAJ3l61Bte0kuVCxi\nX2V9ys3OohWLS7OMl0EDcuPjyI02RHVQi+CdA0UmlS1Sns/hUmTkgiUYOknCK0k0Z7NMHz9OY0tL\niZQokU2SLK7la92Wlh6DRGYaakpxd70bKPmM5caPom3KQYn8V4FNtaLgggcOHhnGaRlgOQVs2TaD\nrx9jZ07Hp3jAkaGokDtr0zf0T9z1xAdwqb4lkqN06T18jO66c2iqFwppkN0Yju+mgtSiJTM5G6ex\nbu0tTAvQ9SXS9Xoqi+Xo6hLTRCYnRdv51q0r161rT8+5c3BHxyO3ANWt0NQIExcs5uRWhqYmaG/w\nCGOUyUlaPR4mMhkcvx81FMLj8VBbmiqi6zrvete72LRpE6Ojo4yMjDA3N0c8HkdRFFKpFKlUirGx\nMUBMuCovL8cwDLZv3044HOazn/3sinYEy7J4+eWX+eLnP0+3YeCVZVp37qR1wsu3viW4YhAeWJQR\nMQAAIABJREFUj12ljo/F81hPDwHvKcpy4EgS2ab38/CH/svNJw2XKS3WhgVfi6sqLSxdTAtxHLHO\nNH7kuqqzG8FqjwsoJe2Gw1iuhvaaEslt5QExEsuWxHoVi8U4e/bsYgEonU5jGAbZTIbY1BSBsTHG\nSspEn9/PREUFtWVlPLV/PwDd3d0oirL0OX7xi3R3d+P1eq9LYlwxHmlpoTYSYXN5OcqlS+KYFQrg\n9ZIoZKlvdGD060tPEtqBWv8EezcrK8bCXsy0UJuOsHnrJhRZESSO3wdWHj2bRPWsbP2RWegRNpaK\nekBusnTuWbZq9Pan6G7Oo5X5GU7OIYYLS5e13qylsHQZsRFP4b8UpdNVvliwMMrL6a3ykgsWkU98\nF/ukirZrE2c2bqR48SKyZWG73fgfemiFOe8dg2wWXnlFjMdaSKglSThxP/LI0pi3a+Ca37VNm4Ty\nYmICysshm0VPp1G9XnKl8dFks0iFAr3PPkt3JoPm94PXizE7Cx0d71jfB0AQE9//Pv7xcaHgWWhN\nNs3FQtg7UcGzlvaQfwDOAB8DJOATwD8C/+5N3K+3BsuNSXR9pfvSDWA5SVEoFG6JtFhIwKPRblwu\njaJ7D9873sfHH5yiqnKWtpL1fjJT4NBomL0fe5+QUAWagYfAtkCfRj4ZA8+EGBVlW5gmOJaJT4lT\n6YkzOj1L7zd/l+7yCbTxOKCA7MFovYsj50ZWBPKGYfHKsZf5fM8X2XZ3N5LLi79SnLje/36Vv/5r\niMcN0pEDHPz+F9bMPmezWZ5//nlOL8yjBBobG3nyySepLTn+Lkc+L6YLgDAVX6hSXm0yy2qT0bVg\ngTTy+cSTL7SZeDxLrS+ybohAQZ9dfFyxKOP3lQOwbauXMn8rhQLMzgovMUCcOG0DuHIUlJtYmcQB\nonpT1rEo/bwaVks0HzQMer78h9Tno2h+GdwVKLJCzbpmLhbu5r9/ch8APfv3M3QuxsyIgllKPL3V\nKtHqJtq7ukR1zLbBskibFfRZn0CxijQ9UqCxWiTTAfdLHO/txSPL4uRTtMTCaFnous7DLS14lVnK\npZRQnIiDQTY1Rg1FcNk0FiRkj0mz349TmmITVxTK69ZhOza2bdMYVdm0cROmZWKaJpZpYZom4XCI\n6b4+ynWxQJk+hfJ8EVmyUHwylT6F8nJ9hbLY5crjxE6Qi8EC55yYnGRzpYFpgmxI+BLCl0NRFFwu\nF8VLFzkbf55Gc5b2qqX2lPORSbZWGdj2PIlLUcKVYXr703Q3ZdGCPoYT01hOPcdOnSSTK3B/h8Tz\ng8OEats4d+hf+WAnPP5oNxvXraNqbJRCMU9P/zR714sWFsOy+M758xypquJ0IIBcLBKfmyPk8WDn\n8wwnk+jAnlAI1TDEF3ahYoFQcNhOqdJgF0rjbSA5LlOhi9Yl+5wLzpeB7IDk4JVthgM6jbu8goCS\nbMACyVgkpJDsZfcv/L3sf9gg28jjUFd08Pkd1KMOyDbIDvL4CJSbwALRAQEkLAMKRQnXnJvk631U\nVCwRItPT07TMz+NVlBVqHj+wOT/OzIEojU1NpfuXnjd35hzaVh0WlVMSvReSdNfl0RQPwzM6eEXy\noUluurV5er/9VTp3P0rPc1+iuz2BVuvFMG16Dh3nieYMyUwZ7Zt2oyguktlL9PxTL3s/8juobrf4\n3UjSZdfDQxK2Lfb5SqTFtap4H/6wyt/9nfiJffe78Bu/IWKQ6xm43mG4c+ORW4HsorERvCMWobYn\n+IezffyXcIyKigqQJPaEQryRSHBGVdm8TLlYKBQIh8M8+eSTqKrKtm3bAMjlcnzqU59ifHycRCJB\nZlnmns/nyefzJJNJvvSlL3Hp0iUikQiVlZWUlZURDAY5deoUuq5DKsVxx6EjHObUUJKR39/Pffft\nw+VSeeIJg/n5A3z2syvVGXt3h3EFN8DpHKZeQU+8ipe/8IWb99Rw+cVv0HFuauzpVZUWU89BsZT1\n1u0B79pbKtaC1R4XK5L2bRtRoofEe7IEsZ9I6/grW4mfPLJIWEiSRGNjI5FIBKNQgFwOx7a5UKrg\ny14vsmVRFolQfP11UqkUbreboaEh2traOHz4MPlSgnP8+HE6OjrW1B50xXhk/34ah4aWVAT5PCk1\nw2TTBJvXbwcULEnl0EiIMwenMYy/W/q8P/CfUFVVPM8399OYi6EFFWxkQEK33EJ9cVcXSGAZecZG\nBjg6lGeWOD7VpLFWY8fWNlxO7ormx7lsGs1tQCGPlE3CnIGNynwsSSZX5PCpl3nt2b9n75YsQVcH\nGEVRWJLVqxMb6yxh6JSLkqww6en3sXfTY/Dww/QMv0p36xxa2Ury49XZSkFSJBKCnV6VLK9F6fHj\n0gp5Rei68PE4fHhloWXzZiEprKm5oae76nctFkNTFOF74XKhBwKMVVXRfs89XDp4UMiodB1HkgSJ\nsSC/zmaRkkloaoL169+Rvg8MDcG3vw3pNJ1tbfQkk3Q3NqJlMqIam8+T1PV3pIJnLaRFu+M4ywOC\nP5Uk6dSbtUNvKXzLWNl8/qZJC88yL4FbNeNcnoBPTOSIRGTs4E5G3Qr+qlYODi1V9vZ+7AoLkKyA\nvxG7bBNUVALiRJ2cTpE203jkNB5fFhuZXDKCxpRIpFWgIkTvwDG6m/JowQDDyWmMYgs9Lx2le0Oe\nB5sgYh+ncX3HCmZ540Y48NR+9myN0t3kXZwicrW2BsdxOHnyJD09PWQyGeKzQwTkNFs3t9NQGWb4\nfC+VlZcvwGffGCY9amA4Kp7NrRjGstFwV5jMcjML93LVBqXTq5CxLqk2Dn6/V4zgWsgLKzuZmPcS\n1y28HhNXnYYU/gX6jhVwZQq0eAqEKwpgFcAuisTRKohre+l+WXIu36HCPJQJxl921t57pqoqex+7\nn95XI+Qms8jpemzJveJ74zjg372PT588gMc3TFZrZIdnhOaWpf5AABSFhGGg7bqLzPl1AMyHoHF7\n6Zjt2EHf/v3EYrEVv4VCoUC4spI9JU+FpHoQLaCKrMu0MN9QIZslkSninzbIeXSIBJAqypEkg7xf\nJVx1F+CAY+NPaqzr/KiQgq66bK0+xaVnn6GYTSO3bSYzMMRYUsITqKSqwsajGVhWEds0sCwDVAV3\noIZCIYdZLCBhY1mC/zNNsG0HXTdQlKVjnslkmJ2dxSiVCRfIjFwmSTYn2sKsmIWiKESjUXx1FkY2\njZmKEbkwR5WTpiYo4+BieCKGlb7Ez+xO4HPLTA8cYl3LDgbn52gmT/eWPL2DQ3TWtvE/X32V08CW\nzZuJplIcPnyYdDpNWVkZXV1dJDSNg6kUfYWC8OpYII9KBIYdDOJvqSYZPIqmzIChg23jyBIoBRJZ\nA3/BB14Jw3A4EE8wrOv0KwqnCx6aSuzgeDKJYYswMV4oCNIElvl+tK9oezEsi2cvDvHC8TiWY1Nd\nKXPv5JIRqh0Dklnx+Uq2uMYh4Dhk5h3y4x6GElVs326hugUhMn8+T6OhCHJFthcJEGQHr+yQjU9C\nXl5GplgggRyZgehKE5tcJIVWZ0MmjZSIweiSJ4oG5M6+TO/hT9O9PoU2qoIj0zuUF+uj7kIzZpj+\n0TyNZWE0JLpzFr1/fJ6u+o6r/i6Vi/DuKQlZkWiVZHAtERqGbdMzcojurTk0f2kRlSSSYyY9//oF\n9m56jF+Iuzh/UcZBYmBQZstWSbilDwygqeqSJGzbtsUq0R2GOzceuRVICi4XbN5kciqmwl37eK7+\nADX2MPK6ddiJBI9t2sTdTz65Jum/3++nsbFxcT03TZNkMkk6nSadTpNKpZBLmV8sFkPXdaampgCI\nRqNkMhkCfj/eVIo5VaWqrJGLs9WYZoyRkQP80R/t4bnnLjeLTMycp2xsgAd23wOmm0+/kGI29CO8\nmzYtbnOlpPm6hqGKX6gsbkJpkc2uGOIgkOyHeMl3oKwdwveu+XlvBKs9LhaT9nwMTfYIwsLSF4tY\nrXe3MDT0LxSLRVRVZevWrVRXVREdHiadyZB2uQiqKpFCAbmsDFwubNsml8sxOjq6eG6Lx+MMDQ0t\n+pmoqsrc3BxNTU1IksTs7Cw/+MEPeOKJJ1AUBUmSrtkWvVgR//a3yY2OItsW5eF5qu+S2dy2HUVW\nMH1NfPa7k0xHxq7pobKcyBlINxKeHCEcbqB9axuKrGDZFoffOMVcNM1Ypp1cRFTsXx8p8MJAnH37\n9mH3yVAzsxSTmRlkV0ycLxwbBwnbtpmYHMQ0TBRZJkgGU5/HXygweW6IxoZGZEnGdmxi8SQ5vcjx\n089x/Lm/5LGODGXjHoinwJGxZIm4laWs3c2fz5yF7wzy4e0JgoEtK47TIvlx0aCL4NLo2RKu1OZo\n2RYnz7zMn/zrFwi3PYSqqhQjp/jVPZVUtgeuuI3X66WpqYnKABRT47dEfqx5m1eeJXfyJeThi9gF\ne6lVdsMGMSbrCrO/byaGX62+uJLvhR0KgccjxvJWVSFPTQm/h5KDvyNJQvK5fj3wDvJ9WJgC9Npr\ni3ep27ax9w//kN5Dh8i99BJyJIIty/h372bvhz709pNYN4i1kBZ5SZIechznIIAkSd0sFSdvCZIk\nvQ/YDyjA/3Yc5y+usM3fAu8vveZ/dBznxO14bWAlaXELZpyrE7VbxUICLklw5Iiopv3Of9571dF4\nV8KKfkhXGfOZMqKGeK7qsiz+tk3or/wLeFQR+JcHQCsnNzCF5hHJjpRL03soQndDFs3lBUnFSI9D\nMYzmcdHdOEnvK/+GnVd5fEuMoM/LfAQaSlNErtTWEI1G+f73v8/w8DCWZZEaP8y/7/aza+e20nGM\nksxMLZIdAD3f3M+D66NImpcN28XHtqnp2j4PN+sPsUAaPfVUL+PjOWxbxuu1eeihJdWGX1tHMjIu\nFAyeShy1jEQKTEfFVHT86zrZvHMDz79WMk4fWtuoQvu8H2rmwDFFW09mCKyiCLRcQWzpBhYXx0Et\nTtJ1TwdoW2D9z698LVuMUjx6VKWq4wl8PvjYxwz6n97PtmX9gcBiT+V7PryHw38h3tMyz8vSMdt3\nzaDxMqmrS0WtrWP2whBvjIbZG3LRG58gGc+iVVSgGwpoTeAVrHsireNvehiqrjTyBlwvRdmsPQJN\n5fD7f8BTf/Ysr5w6iMvloasW5hfa+VygmzoPdz3MAyV23jRN0uk0r3360wzNT3HmtI5t6dTWFAgG\n8hjFArZVJOD3Ijk2RlFHlhxk2UGRwFXUaa01kADVJZMF0rkCmZyBLDkk0zL5fBrbWvKs0NMylj2L\nv9mBAqSmoyiFGYrBVo7pOo6V4weJUf73TJpxoKG1lbm5OSYnJ0kkEng8HnK5HENDQ/hCIZxUilg+\nz4GhIZ7o6FicSJKQJPz33y/6jL9ZoFvLoSkmKB4cXSEZSXFozsveR+/FsG32Hz1K1O/HW15O2utl\nZuNGnjp6FByH7t27wbY5fPIk6WyWMp+Prh07UCRJ+H4kk+z7uZ9DlSSMYpH93/gGgymbWV8YCQfv\nOoeDhQJ94+Pse+IJ/K4CSX8/mk8RX0jHwbIshubmuJhOc2DExTElysaIRnsbjMcS9F9Is8U0afX5\n2KNpgqBZhmmvl00VXRiWxYGhIYbjMUzHJmGpNIU8tITKUSQJZAPZLoqAVbIW+UfbcYjreXTTZCwp\nY9sJOlolbMmHLEvk7BRamQ0U8LrAUPPgE4mQ5oPc1ByEZhAiABmcpWvThlF3lEhtDlmBF+dkqv1B\n7qquxSW76B2fpntLBM3rQqiaJbACBFGpKxvgT5+eIOypYnISYtkCwfMeGo7D1Pw475NlobRZOB4l\npc1yc947BHduPHIrkMR63dFuETgN2axK1H6Cn/tNkHfuFH30AA89tCbZNYjx4hMTE3g8HlwuF+Fw\nmHBp9I+u63R2drJ161ZmZmaYnp4mnU5j2za6ruM4DplEgkw2y7wlM5qfxfC+iqq6CYfH+drXXmFw\ncBC/34/b7UZVVbyqRPe6c2TTec4PDjM1Us1sZhSvd6Xn2GpPDTGm/HICZEWy6wrcMGmxoLRwnFU1\nLSMDk/8mbrt80Piht2QkNqxWX4wjFyLYbhf+TQ/RencLTz/9NC0tLczOztLS0kK1psHZs4SSSeKm\nSbXfT9e993I0FiOdzaLrOrlcjkAgcJm3SSQSwTAMcqX+92g0ukhqAJw7d44TJ5a+/oqiCDm+LF/9\ntiTh9Tu0Ns2hlGWJ52tI9F9gXt7Oi6cH6T9/AbfbjSRJyLKMLMtIksT4+Dh/8zd/s9i24ilfhz/U\nSGXrgxx/6Ws8UpUilRRE2sjoCFOzMV68ECRQvw5d1xefa25ujmeffZaqylYSmQkqygJAANwhbN80\n+LPoeh5Vq2QqaRDPe3ArHhTbomDJuBUTWZYxDZN4PE4oFGJicgLTEPe7rQJmZA53OMNEXKaxvBwk\nmNQL2JUuml0uGqJHwYCQXmDi+Ema1rcil/yakGQ0SSZHGoKVkB+EiQZQ3CC56D3aS3fNGRGP6zKW\nAydPnSGfL/BIi8zzA69ho/Bo8wjn+8q5r+t+gEXFzP3N8PyF40Ll+fpTiyrPBe+qWHKUv/nU07ir\ndmIYBtGhg/zsfSr37th889voOtFTz/CzbQnurdJQNolzVFJx6LGa2PuLv3hF8lHX9Ss+91omw1zP\n98IfCjEbj5MMBmnfvp2LxaIwp3Qc9Pl5UXxZMLaTpBW+D6sJUlmWoRhny7KJd+7yJmJZFgni27WN\nqqqXkU2GLdM/EoeCC/vkSdRsliZNo1JTKG5pQHZPY7/4f/BXtvLQr/4q6tdLbVi7d69w9L4S8Xsl\nYutW39utYi2kxSeBr0qStKDqigO/fKsvLEmSAnwGeByYBI5JkvQ9x3H6l23zBLDBcZwOSZLuAz6P\ncA+/PVjeHnILZpyr20NuFxY6JExTGLBVX9/0ehHLk8TyoIdE6Xwkuwu8NlrNXtvNwagHGmqgugY2\ntoORRHanQEqDY+Igkcvl0XzW4sleymcgLhYdYXR0kQDQUaNjmApOxAVeV8l0UhIL8KVZrPUZzl8Y\n4EzfeRTTps0vMzMb49ef9LK+oQqsSKlCKqOpMt31MXpf/CrgorthDLvgxdDzKJJM4zqJioBMd/Mc\nva8+KwgRSV5hdLliWsIyXM8fAsSC99GPdnHhgvj70UeFWdjise0I0vMND90debRQM/mcKGqncwXO\n58L82if2oKrCnHdwUPgLPf749R3//eE2ktlJkdR7qxcNpSjGSORd+CtvoPesGBNO3CAmyiyDacK3\nvrVk0lyabEt1tUrTdcarlpcLVd2qAkCJaLt6z9+VpK6Gcx/9M15+sSaMak7Rmc3S88YkuyvzJKvK\naW8XMualVqiryNhsG0ZHxe2WFoqGhMu1B5+vD58vhte7Sv0RDrNnmSTO5XIRCoXYvXs3udxBpqY8\nFIsQKIe7SmqSXC7Hrl27OH/+PNPT0+i6LgK+YpF8wWEuG8GngmIpBMwKEkaWlGlQNCziph89k0Oy\nTVyygyLbzKRkfKpNtmDhkm1QTQp6DtUewFLWickgFV6mJkwKjsNQaWbo5OTkisAxGo3S1NTEifl5\nQobBYDzOSCKBLEkULIvBQIDt7e2c/9rXsO0wTx/OohoJLMlP30QQbcJiYyjEM8PDHJ+epn9+Hrei\nYNg26dpaJs+eJZ7LIUkSp0ZHkWWZpGGgBgJkbJtLqRQdHR24JYlYsciBVIonnniCAz/4AbFQiGSy\njnheVCn9O8V1rFDgQDDInj/+c0Eslogsy7Y4fPgw84bEj5INGHd10TcHzw6/SnUG3ve+btKFAvO5\nHBOmSZ/Hs+R67zjCdCocxvi1X2P/3/4t0dpavKX5oJHpcs4ywgweunZ1osgydv441OTRiyaqtw6r\ntpYTZ05iFHO4XRITkguvx2DSyhMzFNob1yFPjYLfWFRZSQFlxQxXuVaFjW0l0sAp7Zfo8T90uo85\nXwHFLxMsA9XvELGTHCnm6Nq8mVw6j7bOxeKTOw4WAQ4PzJNXLPSgRdzj57g9wGQyj0/1sj3ejiLJ\nHNR1+iIR9rW1CeLiRhjudxbu3HjkVlAiLVyKxYMPCtPW+Xk4exZ2bNiwRFoMDi6ZQV0He/bsoa+v\n78oKunCYD3/4w6iqyvbt26mrqxNERSZDPp8nlUqhZ7MUDcCWkNxBZMkiHM5TKEQ4fTq2mAwLODzU\nEiHlypICXjgLfX0m+UgE18wMLl3HpaorkuBoNIrjOJw4cYL+/n68Xi+KolBVVYXb7V5JbmwOAnNg\n3bjSAoSvRSCA+D0vM4uk/mdALV/zc94OLKov2g2InQA1yEVlA9/4xjcWJ4X85V/+JeOHDjH83HMY\nus59DQ14fT7CO3eiNDSgHD2K2+1GlmXC4TBdXV0cPXqURCJBsVhEkiR0Xcc0TZwSAbpaSbHgl7EA\ny7Kwlpv9XQaHWvcs61oGMZ0i8XiRnFTOufQWMpbF8RMnV30nViIWi61oU1p6XS+956bwOOOoks2l\n4XHm0jK4HeSpNy7bvq+vj87OTr4xcZy9W3XKAyqSJDE8HuWEFSOS81BWVcn4+AyGId5PVrd55lwY\nzZ0lXF5AVRyCSZuyWJZ0wsHtUnCZBaZGCnhdFrmCgyybTBcKSAE/eVlCxsE2TRxLHE/bsbGKeeZn\nJgiFQkiIVkYAEx1d0YUIcfxlJFUFCdLDvQQ357ATYtt4LEYFSSoDMgRgyBMDSWJHjYFtz5C4MIeD\nTI2UhKCC40gMehI46UE+sDuG3yMxd/El1tU3YDswfeEi95bleeXiKLYj8d62OcyUxJlj42zfvhMk\niZH+02z353nxXAQHicfbp8jEJXqPzNO5+15A4uK5N+jwpHnh1Umc2Tn23hUhloHDepquTZtQ2trQ\nqqrozhQXY/HV5GNkaoD3tkeIReHw4QRdJWPwtcTwy3El34vifffxtNfLx8NhFEXBHwqRTKXwAGPB\nIO1eb6nVuUjCcRZ9H1bv40Lh9bGNafwRoXwF+OGPnuK5Xgi3igkAt2ObBY+Z5WQTCEKqOTbDjw7l\nKHdEL/q5/CAfvKeax9/diiKL5C+ZmaDn0DH2WpaImWIxaGm54vsSvyvrMmLLsq1bfm+3irVMDzkJ\n7JAkqbz0d+o6D1kr3gVcchxnBECSpK8DHwL6l23zQeArpdc9IklShSRJtY7jzK5+spvC6vaQm8Tt\nbA9Zjrq6pduzszdGWixPEudOD/PaG6Kt4q7drfxSRR71zBn8UoikW0bbvFkEuUoNtmcd+DX0go5a\nUYmcmAUlJyqSjoXDyhPXgtFRMGCTStmYuoGlr/Q0LcSj9L7wZXK5HHUqoEJDfQMjPov15dmlMTzL\noAG5AeFYp23KMTsPjV5R0FjnAuYXSJM+CB8XD5JKVU1JJtd3RPhD5Et97d46CIok+Hr+ECDMuAOB\npQkji7AKqKnj7H20i96BFLmxDUxMmJw6o1KQWvm//tuSbG3XLhEj6rqYJrt9+7U/s5VqBJ+o4ph5\nkrE5DiW2XD1pvxKyI0u3A+sXb+o6/Mu/LOX4NTWCsCgvxV2r+wNXIxS6MmmxFqyWugI88iFhIHj6\nyBHkSASVDn6Q2UDj5o1MXa8VagFzc0tKqZYWzpwB01Tp6tpHU9MBDOMqkuFVWAjSg8EYsZiHRELw\nIYZRoLq6mo9+9KMcOHCAgwcPrvjNW9ZdnBs/zKMdabZtbqe9vR3Ffx6Xb4h0wc89XTs4ffo0uVyO\nom2TyhiMGtX4zBQT+TyO49Cg6VR7Vdy2g6PmiRUrcBQ3tr1yPXGWeVWAqIgZhkGgtpZYPE7SNPFb\nFo4so1ZWsq6tbcX44PtDDh65ghm9Dq/UwbnJw2Smp/G6XLwxNUXOMDBtm6TLRcg0mZ6eXiRJFgLG\n5aRJJBJhujTOUJIkzp49y6lTpzhy5Aj/P3tvHt3GeaZ7/qqAwk6CAPdNEkntkmVJtGQpkhUvkuw4\nluO04ySdtn3TPUmm3Umn3d339D8zc26fM3/MnXPvzMn0msz0cpPuJDeL7Ti2Y1mSbdmyLFsWZS2m\nVu6kuJMgSOwFVM0fHwoLCZIgCUqyfJ9zcAgChUIBqPq+93ve533eUChET49EPC7jdMKZM2kJcVdX\nF319fSQSLn7wah+meC8jg73cGBwjrBVhdjqAVnp6hggExgiFJN5++10cZonQyAgWWabd5+M/nzzJ\nztpaJEkiGIvRu2ULH/zt3/JJayuKoqTeT9PsvNCis3/DCCgXqK+rJ6DJXB+axBexU7milNPXexic\niGEyWQlMJGidqMJhmmRt0Io2pTGmOBg31TJlCiPpcSQSTJnMjCq1yVFRY1Bx0l60DfQEElrqdvXy\nJS4MufBPuTFJOvUlOjZ0ZJOOpqpofo2Ay03AEQF0pGSpTNe4yrjZjMlqJeqxcM2vMumwYKpyMOJP\n8FEEVhVV0VCmMh6Pc9TtFkqbOxR3dDyyFEjJUE6Ps2OHUAcHAsKUf/N3K5FdLvFAW1vepEU+CjrI\nVmQUFRVRWlqKzWwmNBZiUlaxFVdjqWhixYoYJpOK1WolHA6jqmpqLFlTGqCpVBAKNybtXBxygzKJ\nDqjxOOrkZKpDhoGpqSk++ugjTp8+nbXY7evrY/v27ZjNZqxWK52dnbA5GUypC/e0gAxfC99ZmLwm\n7pfcBSWb895fwaGUADA62MWv3vtZ0kvexFe/9CXWXb/OXYODsHWr2HbzZtSDBzn63nt0dnZSV1dH\nV1cXtbW1NCYl82VlZQSDQRRFYeXKlfh8PoLBIImkN1VFRQX19fXouo6mabjdbvbv358iKzRNm3Hf\n+Es8TJXWQrE+jKa70INBhoNV3Kg7hNMKtkQCq9VKIpFA1/XUexh/YSZJYsBkMlGWUZIX6AHZNrty\nWtM0TCYTxXX38vr1Dqy6D0XSiCQqOXUdHt/pQIlEiEajxONxghGN1y+YCUgKgYDOwGgYp03GH9SQ\nRqOooShEIgRDCY6dt1Li1KnwmsBux+K3YZJ1tIQZk6RjlnU+atcxSTqrvHFMso51fIKpDTH3AAAg\nAElEQVSygISEjiTpyOhc7k8QH5kUak71Q2SbGRmN7q4+ep3pMsfJqcksBefkhCCrJ/zisWAoiIRE\nQksIUgSIB2XiCQnHSg0i4BvyQWyYYDCIFgpTapGwhkRZYZlZgwToQZ2ei2Ket4SD1DtkKhJi2Kuz\nxZGQ0KIafedEYFkc8uNW4nTIQDWsqddAkkgoFvrk67h9I4JuliR6Lr9PVfgwXV3d1Ez1U2czAxI2\nfZCVdjE+aFGd7o/78Xq8qc/ac+VjquLvQ+aaJDnfJ7/NrN+9ygbSRoPMl2jaWE9rezvxiQm0DSq/\nc4zRZDJT7nTSNyTWHpFohC6PhzU2Fz3vXuTKlSsU+7rwWoQasndglN9bP0GRQ0aLDNF1fgyQ8Eo+\nvroZ3mybAmSeWD9Gkd1EIjxG5wWx3nEzzlc2w5ttb6IDj68fw2U3oYUn6LgYAUmiiBG+vAmOtb1D\ndU0tAwMDPLFpGIcu0dH6HiQ09IF2KnWVL27XOHpOg5ISHt+p4DD7aW89RV1tHUjCAWBblcpbp4Zp\nrqpDansHPGLcPHXqA2KjF/AqFkieXoMDvRzaNI4daL90mvr6enp7+7DrPg5tgqNtH4Ek8cj6UYrs\nMvHQKO2XWwCw6RMc2gxH2sTa7IubJihymIlHl+6XPStpIUnSM7qu/5skSX9JunofSUSBuq7r//cS\n37sW6M34vw+Yrv3OtU0dUJggYboR5yJR6PIQA+Xlaf+owUHYvMD50Vgknj0LtjawAV/yvoly8TQA\nzVv3cMTuZ29oMuVO7SjyMDThw68W0bSxmWsDLWATJ3YkGkMprQLvCmH4qcfR3C7Q41g9o0yNx5Gl\nOP5oAm+JRiIRZ3RkmOvdITA7kLBSXORg04Z1uItd3Bh4b87jNwiReBzCyZjE6cxOIqbcoSGZ3RQe\nB0L6nVHDHuwBR50wToK8/CE8HhGsZC3Qx89CIoKimNj1hefAvZ6XXoKicSi3ZbUeZ906oTgLheDs\n2flJixlqhIk6tOANHEUeDn7tTxZWexZMshJmO9iEZGdyEn76U0GAgSjX+/rXs7m7+eDxQFfX4kiL\nXEiRJAcOwH/+z4Ld3rEDvvjF/HfS1ZW+v2oVLb82jlXhW996NO+ksxGkOxxHee21TjRNBOm7dqWD\n9FzZR5PJhLViG+8NjeFcdzc3OjRixXv57ZUivn5/KV63g/q6erp7ugnH4KMBOxu37sI31IHN2Y3d\nAhUr6qkunQI1QI0EE+Ym5PWHqL/YwfDwcCoYPH36NIFAgERCmJAqikJ1dbUI9KqrKSoq4uDBg2ia\nNuOmayrlJhu6bmXSVkO1s5qyxx6j+/JlpkZH8Y+NETWZkO12ypMS8kySZDphkut5VVWJRqNEIhGC\nwVhqWDWZsrpskUgkUjXwWDzELR6uj44TiiUDjaRUWdOGicfFtdrW1s/atdUMhMO443HMssypvj4S\nui7KOex2miIRWlpacmbsNM3BPx2PUlHUz+oGE5GEid/6FB7faSHU2ZlSsRhB6qTkhtgUvRV+nDaZ\nYDBIQLXR0efH7ZCJqhq9PjdjMbFungrGea/DzZXg8Iz3Pn16gM5OCIcTmEyifFfEVxJgweGYYkOt\nlWJbdqYyU1nT06twpV94qui6mLIiET+SuoJNpjjFdjOd0y7MiSXMa7cTPhPxyFKQVFqgJ1AUUQXy\n+usimXb+gsS2piYh+evsXFCntPkUdDBTkeF2e+i92ANRnSKzjcp1n2fzvcXY7aKsZN++fXR2djIy\nMoKu61i0Cba7zyLpCmHVxLnIBrZs8aJOTjIWChHXNOIOB/Hi4qyFsd1ux+FwzDieUCjE1atX2bhx\nI5IkievH6CCyAKVFJmkRCCA6Twy8kfxiiqHmFncTsHgYHR2l9VIrCvegmYr52u7drD12LN2r3W4X\nc+nmzSiQ1aL2B0kfKkMy39jYmBqTGxsb6ejoYDK5H4/HQ3Nzc2pb43fcO72XfC4EukSXlVgJUALW\nYrhaBJ4V8J3nUptJksTIyMiMlxskRmlpKd/5zndSv388Hk/dz7zpus7o6OgM4sPYT3FxMQ888ECS\nVPl8FrkSi8W4fu0sjAxytTPEVCDGpOrA7C3HA8TjRbx5tZ2HN8ewmyExNYUUjRKI6rx+Xga5iAmz\nTFCJ4jTLaDpoCYl4XHxvwbBG27gQiXWMinnFbDajF6XNJ6eCca6HK/BdS/oRJTYmzSl1AoHTrA0H\nBRku6QyOgpaII0uiVPXysAmTrLPJryJLolxVlnQ0TfyVJRialLGadSbDosGX2axhCceZDETQNU0Q\n5kkiRFXT8fVUQCy2DZIkERPxdSya3sYf84GqoiUVN6omFCERyQQmGbQEfv84ekZsPjlqpuu6NENF\nKscm0WLpfQcmQsgZ1+/kkJmuS0uf3ySHqAes3FzOoM/HjeAwkt2PDpgUE65iCz2tQqk2cuMGVlWF\n5DJPiUziNmsQAxkIjou4RUlolJjBFhEZT49ZA1UsuENjYwDYEho2Mzgiop1aaXIbgPCoiCEcCQ2H\nAsXRIUpCPQQjk5QrGmgQHuiDmIpL0UABuwtKqlWQ/VQmt4kM32A43Jb1eTvVINbwdbj4IQy9DID/\nxg1q42qKsABQo5NUKYK0igz1MhS6SmRqkmJNo1iB0qjoJlWdcdyRQbHucGsabjOURbqRZYm1nqTC\nogDVqnMpLYzZoIiMIKGAyHef04sFc77ur//6r1P377//fu6///7591wgpUVmeUghlRaKAmVlQuY5\ntISwyFjTNY58iOeTE+Ib9XpRnn2Wg1ZrlmQ/5riXF67Y+Pr9pZhkE44iD/7QJFYz9Ew6adq4KdXW\na2IqgmOF6Dynmk8QkK2iO6NfJ+Ea5XrbdcZ8Fq5F1zAZ2MDnP/957v3c59L9lK/9nfBwQBNdJXQt\n675WVgK6xkB4lJGYyFhWVGrg0FPya83thqqHZr7WNQKOCWFUFRWDBIlwirTIxx+ipEQY7aaUFloi\n3ZPdWgbFwhws2QmOFSuyS1vNZrj7btGisLNTBJDeNFGcE1lqBP+VdBuwWD/Y8syg6roIFAAc4qBG\nRuDf/z21FmTDBnjySXGMC4FRDh0IiJKYgnn4KIpgUTo6hDxlITBO8KIi+sMejLXw9u0LV8krisKz\nzz6KETvdd5/ovpX5fD7ZR4AHDBOp9k50rYg320KMhxQq6kUrLk9lI4cv9vN4MzQ2rRamrONn8Adj\nnOzo5eAf72E8KBauVqs1RVB0d3djNpuJx+OsXLmSNcnMuhFIzrrIiIzCNaG6aK5/Ajxbs57++7//\n+xlB44cffkggEEDTNKxWK5qmEQwGU4GgxWKhoaEhFRQWFxeza9cuhoaGaG+fwOXS0XWNujqwWvUU\nyVFUVMSaNWuysmrXrl1LGsSKbXRdx+VSiMV0YjFQVRlVtVLR0MDk2BihcJiIyURvcTFKSQnrmpqQ\nZRldF6ao00kWWZbBVkrAZMNWswsbUFST4EiHyLbd6DERisBE1AYWD7Iso1k8vH4hyBe2xHGbdZrq\nPRy/EmTP6hgjQQu2YnFBTIXiHLtix1PfSC7E41qKwLHbZ5bAa5pGVCplKjRJkSN9URqfIRjWmIja\n0HUxT0mSQYjqWFy1nJkcYLc5jKppHO/q4nhXV4rIuUNw58cjS0EGaQHQ3AwnT4q16zvvwJbPN2E6\nf14YzfX1pQzmCoHMMfH69U4k3UFR5BwWSxH11Y3U7S7Gas0uzTt69Ch9fX04rGa2eq9jN5vQdRNX\nondjK7Kzb5+IK05cvYpV10UJ1oa0aWHmWOd0OlPE7qVLlxgfH2dkZIS+vj7q6+vFmJwiLWLCeFye\nf+LKLA8JBjTo+414PQgfC/Otvbau94zQf6kVXddxyhG+XL6GNe9lJILWrIHHHxelhtMw2zz253/+\n54CoRS8pKSEYDGKxWFiXbB8JuUssc0JLwNBbMPp+uptV8Voo3w2fnAI1KLpHJOPnTMVOJiRJIhqN\nsmbNGux5jGf33nvvDDWkAeO8+Xxmve8MCJ/f3/3udzn3k5hcR8vH77G5WMMfVZgIx4kmXNy9YytS\nTQ3xRIJLfR9yoCZMmdeFruuEQiGmgion+2xs3r5FlDUNn+fgpghlHhcbN24EHSaCUY5fL+Y//PFX\nMP/kJ0Jv19yMfvfdaJrGpY9LKC09R7HDDLooW0l5yIQSBCyV6OiMq0M4bTKaYkXXdSKqmHymQnGu\nTJXhMgfpC4nHLBYLSlUDnTfaU+uX0wMKJklnRW00qVnQUcwmJEknHhf+XZ8Mm5EknTUTKnIijhSN\nYNYTgAlNl5BkiXafDclqYTCUAEloH8xmE9jFOSkBYzErzkg1bWN+4X+Q3C4QiDARSSQHXF0Ynzvs\nqQHYr1oYj3lS+0mfMPrMx8jjMUkHtxtTkY4UuC7OWbOXoFYsGqIBYyELcTW9F1WTCcfT+5A1GUkS\nCl0J0JL6FlUTfyVEG2dIXxIm08xJZfqsopgy/+qgxsWbGDDJwidOmRbw5pidzIpRg5RmKHIlpJRM\nXluf9nf683Mcd/tQjB/8diLHxovDrEsWXdd/lLx7TNf1rJR40vxqqbgB1Gf8X4/IXMy1TV3ysRnI\nDBLyhskkBsxY7LZUWoDwtRgZEUqLxUDXxZqufLiVbcOHkcoRs/HTT4PLhQIzJPuZiy3NXsR/+yDI\n/rstqd7YMNNn4MivWqlxjBMM6ly/fp2xsTHCMRHIb9/3EE888QTeaSv2LA+HaVftxFQER/U9xBPQ\neu4ENsVKcTE4Sqdts2IfVNw343M7GlT80gncrngGaREBxZ1qDTYfjAX65GTSPXziAqjJDEb5HpAk\npqbSqoNk+XwWtm8XpAXAxx8Lk+S84WoQPh26BoF2KM6TtFAn0sfpXEVvL/zsZ2lebscOYQy6mLL3\nkpL0fZ9vwd2p5kZTkyAtxsfzY3hAnOAZfhYtZ5PyQEl894tBUZH4XMPD4nCmrzfyyT4a22VeW/ue\n+P6MIHHHjj/H6YR3OpNGR6EdOBJ9HNzXiDL0Ogf2/15WFrOxsZHh4WECgQAul4vGZPvCvAJJNWPi\nsMw0RMoVNHq9XqamppAkiepq0TlmOmmyMrkAMgLCRx55BE3T+NGPTlBWZsVigW3b0gv12ciVQCAw\ngzRJJBKUloa4cQNMJgdO50527kwrxcvLy/nud7+b9ZpM8iWXUqSsrIznnnsuixzRdZ0f/vCHOV+X\nSCS4ONyJV1Ipq7yXVaUyHw1PUNtQAnICDQVH00r+1z+6P4t0ydxPOOwgGBQZlXXrBBmduZ3X6+WZ\nZ57hvdf+X7Y1TlDitKAnA7WRsSk+6Lex7d5tmC9ezFKRhMN2XK5qMK3ksrWDsDvO1q1b2bdjB46G\nBpoPHODnS2jBfbvgMxGPLAVyMpRLZjDNZti3D159VZDu5wNNpIbD9vaCkhYgxroHHniUgQG4r+Q8\nf7RhikvRDrTPr4LKkhnErqHOqIp/gNMszufu8CqGgg5KS72pcaz13/6N8d5erJPpKqDpY13muLVh\nwwZaWlqIRCK0t7djsVjYuXMnmDNkE/EgWDImsllgs4kQMZEA88T74EpmJ0p3io4htxCXL1/m5d++\nya4SHTkW43GC1KeUaxZ4+GExAc7T0WO+eez73585Z+XVcjYyAr0vQliUFCArUH0QvPdAIKMNs8+X\nMm+bz0NlXpIkiWXbTzQKXV3E+/pYY3Pz3JZdHO3u5kQ8TlFjYyoDZFYUlIa9/O76FTavLsNMgqtt\nXSQs5VStTXdlS5Q9xLGuK2w2lyH716ZKYf/wz5Lf7cmTQp5YUZEq6dq2bZvwgaoRPlCRSCSt4Gy3\ns36L8BQ4kyxXXbtGnKfGNie67WzdKVSejiKh8ly5QiQ+YrFYklyJU7VKjA9Od3eKRDdUTcY2xfUr\nIRrBHG2lSIuAZMVhER2sQnY7U4oTs+GZ4OjBZTel9rPpnntAB99UhN2r97Jz78P84z/+I6OjoyQQ\nc2OQNvxyL67k+yt2O6ubxffgm4qwbeUeduw5mPq95lOCLuixH/0I3eeDpnXwxBOphy/90z8xOjqa\n+j8cOsuUkp6PDVItnCz7HUT8Py6nk+J2mx1d11OthHviYpsBLZQiaOxWOyC2kYBu1cZgfDPR0FmG\nBgcgruJQzOiSTMTpAIsddOiNi+qBEU3s22axUrtyY3Kv4jM6L09QnSiBuBnuexqAD37zIhM+ER8a\nQ0YofoVRXayLbWYr1fXr6Lt6lWgkIsxl41ZB5GjJtbMktgNjHazTr9pwlsHXdpswGI3/55WPZnz/\nC0E+eda/BbZNe+xvgEUuCVI4A6yRJGkV0A98Dfj9adv8Fvge8N8lSdoFTBS8ftRmE6TFEpQWRu20\nrusFJy2qqoSZ1tSUKFVYaFdWnw/k7k42XH6RkkZdTGhPPz3ngnD6YuvzX/4+Le8f5Z3O7LZDmT4D\n+5/8Pv/wg7/n+OHjKFIMd6mDkqoqnvnen9Dc3JyzHdaMjhJJZBIiFy7Ary60sn/DOBs25N4mF1L7\nXjmS7gueCM9v6pgBY4GuaeCf0PGMnEx+QcVQsgVIqywgN2lRXi7aPff2CtLigQcWQBaYrOCoF6Ue\ngQWoDwyVBdA2tIpfvCxUESDaYN933+LNzjON55eFtDh6VNzv6MiPtBgeTl27sZpVXHxbPLx2bdqn\nYzFobBS77usTsUohmjDkRXboOvT8CvyXYPIKStH5GRmxQ4cOpZzLNU3LP5CMZbZ8mRm05wr2pkuG\ngbxIkwcfPMDf/m0r8fg4VVXW1Pk2V+CYizTxeDxMTk5SUgKq6iEeh2vXYNMmiEYjNDTMJB8z95M5\n7hjmcqtXr875Pa1du5ahoaGcGTqpspG9+/bxSB5kVS4UFW1GkkQ3m+rqbIVSJBLh7rvvpq6ujif/\n6H+j5f2jXBsSY+2F8RCtHWNUrBaZzulKm40ba/D77YRC4IvU8/Uv7eP+b9xi2fry4s6ORxaLaUoL\nEEThe+8J0uL4R07urqjGNDwgSIsHHyzo2wcC8G//JhSh2/s/orzUxJ/s3Ir5r/4iZymKoig8/837\n6Xr3HL4JBz61hKCrmfvuasoax57/9rc5+s//TKfPh2q3o7hcM8a66ePWpk2b+Pjjj4nFYvT09CRJ\ni4zWv/FAXqSFJIl4SwsO4gwkJxZrGVQvwFtqGXDp0iV+/etfo8VN4Jtks03Bq6hCor1ypVhk5dkh\nZj7kS9CnoOswfgYGjghFC4C9GuqfBFuZ+H96EJEkLRaiYpzvmAu6n9deo/PNN1Hb21GAhhUrONDU\nhHLPPRx4/nla/+VfxLmXIVuNx+Os2bST554XHfCml+JM3ybnMXk8grTIKPmbXkKcS8EJZJWryro6\nj8pTzOGeEg8DQz7euuZKKQaPXR5m/4Ywdot4HhDbXLLhiakwOMKxjhj7t2him9paqKhgoL2Xt646\n8dY3JfczktqP1+PFJJvwB6J82FfBwaceRVEU1q1bx/DwcGr+La9bx1vXfKnXlZeVY7FY8AeinBmo\n5uBTjy9fu866OpHIjsWy4tDNmzdnqW9C1kqicUHsxONxqiqFd053TzfhKEgOkeiJJdLbVFdVZ20j\nO0W7xbie3qampi57P856igMxRvvCxCrAblGoKq+G+hV0Dw5gNpuZCsbRk/51EV0QUlW1KymtTC9M\nJqYiNKzdwrqBpOR6dTO4XGzpD85QFU2F4oQ1sZ/K2pWUVzcxEdBSx5RIvlco47graldmHXfcJf4f\nDqS3WSrm8rTYDXwOKJck6S9I58KLECVAS4Ku63FJkr4HvJHc3z/run5ZkqT/Ofn8j3Rd/50kSY9K\nktQGBIE/XOr7zoDdLlLpSyAtJEnCYrEQjUYLWh4CM804G3Orj2fFjTMDbP7kvyPrCUpKTcLEIJkx\nzRe5DBQzMTg4yCuvvMJEaJKIZTv+KFRV3s3//n8+nLPmNHO/0ztKTCdEPv4Yiuuf550bR7Gt7kSZ\nzE2azLnvtuvIiTCaQ8HRdN/cpo4ZyFQVBAev4oknGday3akSGSPJbzZDTU3u/WzfLkiLQACuXxeZ\n1rxR1CRIi8gIxPxgcc//mqQJ540hGz99vxJdF0TJoUMikF0KpscbBUVlpSgkXohhXIafxaXgSozL\nr7l5aYfS1AQffCAIq66uBf5mS4EkQe0hCPUJtczAGyjOlQsLHGeDobSQZFAWJxnOlzQZGlK4557n\n6eg4yooVnZSU5G+Emos0cbvB621kbAxGR+HGjSjr1uUmPxabaStUhi4X3G7RzUZRxlGU2fc9fazd\ncyi79ny60mb16kaCQfjooyg2Wyl+/wGhCrvDmod8ZuKRxSIHaWEyia5XL78sQpxrShMbGID+frEY\nmmNuXgh8PkFYjI+Da2qAJmsfGzaAfG/z7N4ZkRGUoSOitM3shDXP8bDimrGZsmpV2lj2iSdyDsTT\nx62SkhIsFgs9PT00Njby0ksv8YdfO5gOdhfQ9rTIFadGepG4mhDjZv2X8yotWS60trbywgsvoE1O\nYr52jU33FuG1SxCPwcGDsGvXrbv41YDorGIYlUqSUKRWPJCKl4A5g4gFkySzoCD7SSRQzp7l0WvX\nRHmS0SVq3TrRDq68HAXyIkgWRaJ4PHDjRs7vaD4F5/R956PyVPV7OTFow1pRmiI/iut38eqVq3gd\nKvvrtqOpGic+GsE6EsJkHgZZplir49VrcbyrPOyvehBNUjgxeH7O/fS222bE8NPnX5PJlNfrlgUG\nUTE+nmp7musYPZWNHLs8zINrA5R5XSkCqKOnn8OfgLchTf4sepvzcbyhIPj78SglHD4/xeMPltB4\n4ABIEsP+CUbHA1lk03RCCjISvfc9DL/8ZfrzuVw5Y59cxFZjU+OSP9tSMZfSwkI6IMiMcCeBryz5\nnQFd118HXp/22I+m/f+9QrzXrDBq5JZoWGa1WolGo8tSHmJgcHCBpIXPBz/9KeZEFLMZnE9/eeGs\nxxxQVZXjx49z6tQpNE1DkqCpyYuiPIbX28ic3a+SmIsQGRgQmW6TSeELX3p0YaUVmfuuGxRGnM4V\n0JT/RJaeW3USg+9BGWCygTed1DOUFjU1s/tDbNokTNFiMWHIuaAFsKsJeEvcD7Rnvfds0APddHdD\ny7UV6LqMosBTTwn1wVLhdIossapO66pSCEiSOD8vXMjfMC5JWuhOFx+2iaDC7YbVq5d2KCtXpqXB\nHR03kbQAUStd/3vQ+WMh9+59AZq+nZaALxaxZPCjuLPaA2eiUEFjR4e4bteseZT/+B+zTe1mw3yk\nSVtbL8ePq8TjCprWwLe+lTtwWWymrVAZuumYnISxMdHNxuU6isu1tGOaThpVVSl4vQ1EowcYGlI4\ndQr27FnUod7O+GzEI4tFRveQTNx9N5w4IWLT9waaWGt6D5NJFxfoQp29c2B4WBAWRgOw+2wfsXEj\nSLI0O3OsqUJNpqlizF/xJOQgLADhbG24kff2zjoQ5xq3XnvtNT766CNu3LjBsXc+4BGjImYBpEWj\n8y1M0WFBhlfsA0ftvK9ZLnzyySe8+MILaF1dmHt6+MbmzZTbesEVg7t3wPbP3bJjY/Ia9L2c/m4t\nbqj7MrhWzdzWbgej3KLgmY8CQNehtRXefDP7+GprBTE0rbQqnzlzUfOqEYD6/XPGQoV8//u/pM6Y\n/3bv3s2B++5D+egj+OAD7t98L0c7OoT6qawMZcMGdt99d9Y8Nut+FjH/zve6ZYFBWsRiIomW9IXJ\ndYzl5YfojvlwlHk4kex459zy5zzclE70LHob71M8bD1Nr38Y1WZDcbnY8YX/gHNVOe90JcmmskN0\nB3w0bU8nkWaUHWcmejNK7RgfhxUrZv3uc+1nqZ8Nji3pp5nL0+Id4B1Jkv5V1/XuJb3L7Qyjg8gS\nlBaQNuMsNGnhcqVbb85lxqmqalYfYk3TsF+9Rry/HDAR2vcI8pbCtedqa2vjtddew5cc1GVZZs+e\nPTQ17eO//TcFXYdLl+De6f7rC8BHydInScq7S1tuWLyCtIiNz79tBtxu8d5uUzfaVJ8gLUp3irIN\nBM9l/Ca5SkNSb28RnUNaWoS0fXJyAaUL9upU69N8SAstMkFb6wT9AzChrsRuhz/4A6F2KwQkScyl\nw8PLFG80NQnSIhoVWYa5vtgMP4vx4lUMDAg2fDEGnNNhsYjvrLtbxPc3Ha5VUH4fDL8L4SEYPAo1\nX1jaPo3ykBx+FoWG4aVaVZUfYWFgvuDqkUfSSYIjR+CrX81d6rRY8qVQpE0mjO/CZFJ4+ulHszoM\nFeqYVBV++EMYG4O33077Ztwp+MzEI4tFDqUFiHHw/vvhxReh37yCviELK6tj4qRcImlx44YwdzZC\np/vuCbPz3EWkOOIEdM+iCux/HSLJDjvl96VakeeE1SpqEIeGRAZjAXj44YcZGBigr6+P0y2t3G0d\nEtLseJ5tTwNd1FpOMQiMh2vFsd4iXLx4kRf//d/RL19GCQT4/bvuorG0FDZWQk0ElDwyRMsBTRWl\nIGMZdeolm6Hmi7MblRpBxODg7UdadHWJEtWMNuF4vcKMbOPGxdfVLgYGaaFpImgsUMnPXJgx16gq\nnD4N//APqQtdMZl49OBB8Z3MEljeTvPvolCaYZ43Pp5lZntTjjEWgzfegM4WWLVK3LZuFWZ0OcpX\nF6QBLSlJE8Hj6TXRzfvu/2xJr84ndfdPkiQ9pev6BIAkSV7g57quP7ykd75dUEClBRS2ewiIc6uy\nUiycZjPjVFWVIz/4AXvHxnDbbIKVPXeO0REfPx22U7T7L9hwYFdBjicYDHL48GEuXryYeqyuro5D\nhw5RWSlKEUpKRBa+tXXxpEU4DMZbzBX/5AVjkaYGhPu3KT9jOpNJkAv12klxeshmKEt/oL6+tAPw\nfL5m27cL0kLXRee5+/KNfyRZBHUTrRDoEKacs2TJVRXe+k031qSXoeZYxf/0jcIvXpaVtMhUArW3\nz01ajIyk+mh+ElgFiCB9qSUwmYfS3S3eZkFEU6FQ8Xnxm4f6YPRDobopXoJcJruCXL0AACAASURB\nVEVazF/PvRSEw+mYr6nAfnUbNwrlUmsrXL4s/hYgYbysaEt2HLPbF1yZlzcUBb70JfjXfxWm4L/9\nLXzzm3demQh3ejyyWMxCWoC4Pt59F0ZHTVyYXEVdxTVMbW1ZsueForMTfv5zUuV4+/fDXvN5SLYn\nZseO3C/0XRBtwwGcK6Hy/vnfrL5ekBY3brCQ2iez2cxXv/pVfvSjHxEMBrl8rQOXy0VRPkqLRBT6\nfoNF0UnoCufHv8wByTTT5f8m4OKFC7z4d3+H3taGAnzjrrtoWLMGvvxlUDph8JhIaiQiQgl6sxAe\ngJ4XIJosmzVZBVnh2TL/a0tKREBbcLnmIjE8LMiK69fTjzmdor6qeY4yp+XE9DKam0BapJBICBO2\nd95Jy6hAqE0eeqggim1VVTl6tIXOzhCqKqMoGg0NDg4caL65qopcyPRTGx8vuHHxnOjvhxdeEBkI\nEIn1Q4dE4FMImEzi+vP5skiLTwvyIS3KjQABQNf1cUmSKud6wacKt7nSAkS2sqND1HLnUom1HD3K\n3vFxQVhoWsq506qa2WEt5me6zhdWLe0YdF3n3LlzHDlyJOV6a7Vaeeihh7jnnntEO0FEDLRxI7z/\nviidWOxi79y5tHnkbPFP3rBmDEAxH9jzP32rSoYoDVwnHAE829Kt00iXhkiSiKvmQk2NIJ+GhkSJ\nyN69C4gXXasFaREPi0Ahh0Q1HBZBpH2si2ob2J0Wvv6VaoqWYaFtzJ0+35Li3twoKkp/Ue3twrl0\nNiRLQ+Jx+Hh8JShLN+DMRFOTyFqDuP62bp17+4JDNgnp9PUfJoPol2HNc7NLqedCIpYh3V1e0qKr\nK03mFbAaLYVHHxWLplAIfvc7kYRYiJrjZsAIyNrbQ7zxhoymaWzd6iCRaEZeppr4FStg50748EMx\nNp0+LUrc7zDc2fHIYmF4BmgzSQtDbfHrX8Ogs4m+vmusNE0JNnYRTspXrsCvfiViEUmCL34R7mnW\n4e+S2fbS0twXfnQM+l8V980OMbbNQsBnob4ezpwRAcHQ0IKYv+LiYp566il+8pOfEI2baW1tZVv5\ndub1Ve5/HWITWCzQETpAMF5GOFwwG5C8ceHkSV76r/8V3edDkWX+YMsWVj3yiFg4Kkr2oj82Afaq\n2XdWKOgajLwPQ2+nSTLnClHSmO/csqxBxAIwOSkm+XPn0pOWosDu3aLGrhAO3IvFshqIzQJdF+uH\nt9/OXtCWlwvz3vXrC/JbqarKD35whLGxvdhs6YxkX5+f1tYjPP/8wVtLXJSUiIFT027ewl7TRMeY\nt99OtzNdtUqQk0vK2uaA1/upJS3yoawTkiSlaKaku7Y269afNhhKC1XN6lu7UCyX0gLSvhaJhIgz\n0HVRZ9XbCxcvEjp6FHdXl0jhf/hhaoCbVLzEKjbhDHRmeWMsFGNjY/z4xz/m5ZdfThEWGzZs4Lvf\n/S47d+5MERYGMgnBy5cX/n66ni4NmS3+WRAy5fCxhQ3+TUWiu144IkN5ds2oQVpUVKS5r9mQ2YLT\n58vyj5wfme3VptpmPO33iwxrTw+UKF2UlMC2vSsoKl6eNKthUKqqomyp4DAMKW7cmJtMTJaGDEw6\nmTALOclSDTgzUVOT/l1vSYkIiHO35ovifjwIfS+lg6uFYJ52p4WEUQ5hNs8tlFksnE6xUAJBXLz2\n2uK+kuWCEZC9++4GOjoeJBC4n1DoQQYHN/CDHxxBNdjYZcBDD6Vj3Tff/FTGJPPhzo5HFouUp0Ui\n58WwaZOII8a9q+ntTYY6xoW6AJw7B7/4hYhFZBmefDJZutnRkc4M7tgxc2GjxYWPRSIZH9X/nujC\nlQ8yJei9vQs+5lWrVrF//35imoVIJMKZD99B0+Y4ZfyXwXcOAN3VxI2IyJoE8qwqKQh0nfO//CUv\n/af/hO7zYTGZeHrPHlb92Z+JGjljQZdJEiwwtlkUYn7o/IlQd+hJc9Kqh6Dxmwsjw41BKh6/yV9s\nEpGIGCD/5m+EosAgTpqb4fvfFwv0W0lYgMi+GLH1cpMWug5Xr4oawxdeSE8cJSXCAPe552DDhoKR\nS0ePtjA+nk1YAFitbsbH93L0aEtB3mfRMNQIkB7XlhN+P/z4x+KcNNRk+/fDs88WnrCAmUajnyLk\no7T4X4ATkiS9m/x/H/Cd5Tukmwx7Rt1dJLLolJ1BWmQqLWb4TCgKjoYGmuczldF1sVibmACfj/qe\nCdZcm8AWmSD+Ax+4JrIIFvny5RnlLXpRMb3KJlBlKtzqomTCiUSCkydP8u6776Za1RQXF/Poo4+y\nfv36WV9XU7O0EpH29vSYmSv+WTAsmUqLBUTxMR/l1la6gb6pTcRlT+qCicfTJbb5Lsy2bBEKxHhc\nqC1ydGvMDaUYbOWig0igHSo/n3pqeFjUFU9OglWeZEWVjw3rQS5ZlefOF47pCYCCZ7mbmgTjrOsi\npb5x48xtdB26utB1uBxeBV6JkpLCliPIsiC6r1wRMfktSwh5tkCgTUirp9ph9AMo372wfczT7rSQ\nMNZCK1dmt/YsJIwSkUuXbo8ykUyp6/nz1+nq2kV5+RSy7MJoblFZmQ7IHn10eSQQFosoE/nnf1a5\nfr2F738/NP+LPl24s+ORxULKkF/qiTSJYTwtCbXFLwa9BMwl9PVNsKqtTWSU88QHH8Dhw+K+ogg/\nGaOxRyrLoCjC/XM6Bg5DOFnfWrEXihbglOz1ColDKCRIi507839tErt37+aD3kaiwx8zMdLH8ePH\neTBX21c1ADdeEfdNNhLVX8JoVLMsBH0uBIOc+7u/4+XXX0cHLCYTf/D1r7Py2WdnLqSzEjLLXGox\ncRFuvCbKUACspYJ8Wow56fQgomhmN6tlQTwuVDvvvpsqLQWEguChh4Si4HaBLItAenx8eUmLri6x\nWM4gBDWni8Tn9qHetZ04ZuLJJcdibkY+OPN25EgIv9+NppG6bdkiYkmr1U1n520wb3m94rtfbub/\nk0/g1VfTa7jSUsEGz9aOsBAwrr9IhFsiIVsC5iUtdF0/LElSM7AL0IHndV0fXfYju1nITJGHw4te\ngU0vD5nhM5GEv6+PI62tHHzuOZRAQKzsk+RE6v7EhDAiTKJEg7oBsWiK9AHT5ntNlkWwYLOJm8tF\nxFtLpCXZbqhi4SuHnp4eXnnlFUZGhEGCJEns2LGDhx56KKsdYC5IklhUnDy5uBKRzPinIJJ8k13U\nWyaiC8tGjJzCbhUZmd7wHiYm0v4QAwNp3ihf0sJuF2T1xYtioRUOZ3Nmc8LVJEiLUF+qdrWnB372\ns/RYt3dbFxtXJBfWzuWrwZseb8xXGrNgrFgh0vTxuFgB5yItRkchGGRqCroRn7UQBpzT0dgoSItA\nYNFq6sKg5osQ7BXn7+AxYdRpX4BBQmZAu4zlIZnx1XKUhmTii18U8ZZRJtLQIFQYNxvTpa59fRAI\n3E1PTwRVbae2tomiIlNyvbH8AVltrcqNG0fo7t6L2bwMWZpbiDs+HlksppMWOUK79euhukZi3NNE\nX18LtW3dKKo6L7Oo63D8uChvBxFifOMbGfOe3y+ytCAcp6dPahOtMHZG3HfWixaYC/psklBbXLu2\nYDPO9C4k7rl3Hx+/eRU1oPLuu+9SW1vLusxuJLouWnbGk9dn7WM41XTgclMEAVeu8PEPf8hvz50T\nhIXdztN/9VesmK1M0mQXHl2JWLaarpBIRKD/d4I0N+BthuqH8/YHm4HpQcRySPIyoeuo585x/Yc/\nIdA7jqZJyLKObU0tG/70j1CW2m6swNB1Ef7oTg/6wDjxPh+hkYWTA3PdlLEBqlvfpGioLUUcxGQb\n3XV76K25l8RRCxxdns83NibPsBHM7DaoqreBGdMsbU8LhmhUBC7nz6cfu+ce0aHGssjrKl9M9+y4\nE0gLSZI2JPuUNyOCg/7kUyskSVqh6/rZm3KEy43MCXYJvhbTy0NSPhOyLAyHgkGIRHCHw+wNBmk5\nd45dqTTF3JBlsLmtjMQ93CgqYfVuj2BgS0rA48GxbRv+Dz7AnUEmTAyIv1PxCDXb8k3pQyQS4dix\nY5w5cyb1WEVFBY8//jh1C2hBYZAWILKh+dZWT0yI2AQE8zpf2UVekCShtggP5E9axIPgO4vNDuOx\n1QQSVVmkhVEaAgvz6Nm+XZAW8bhokpG3CqVotciw6xoEurgysJ5f/zpNnOzfD3tWdSONI/rI25eP\npS3JWPMuSwLAbBYSh7Y2QVrkmDDibW10Xu/jyhWVY8U3CPS/xV13OVDVwpo4ZSo32ttvIWlhsooa\n8PZ/EYuSnhdg9XfyDxqN8142g7nwBhCG0uD48RAXLsjIssaaNQ527lw+Uy2jTORXv0qXiXz1q8vy\nVnPi6NEWhob2Istuxsehr09GdBWzoWkr8PkG2LQpPXYud0B29GgLpaV7cTrdLIPF0i3BZyYeWSwy\nlRV6HHK4NkiSsAg6cmE1NQMt9HXFaejpmVOeputCXfHhh+J/pxOeeUb4bKVgOEzDTAOq6LggAiDZ\nyvkraf+NhaC+XgQGPp9gDxaRXFLsJWzetJmWs2eR0HnxxRf5zne+Q6nRKcD3sWjdCaIDRslmnBn8\n4rIqLSIROHyYs6+/zitXrwrCorqap//6r1kxV79tSRLKucTw8pSHBLuh96U06W12QO3j4J5daZsX\nlj2IyEBnJ/HDhzn94gnC4RWYzQ2E7KV0NjxEn70G70snef75lVnzlK6LRXQhyIHF3IwF/NprHmr6\nQTX7OOkvzNfhCI3S0PkW5SOXAAgDCVnhRt299NTvIa7km0mbHZIkuFCzOfeto0MjGBRrG+OWGSYo\nym1Q8TdL29OCoLdXtHQyzn2HAx5/XDDLNwPTSYtCtRe8CZhLafEXwLeB/wsRJEzHAuny2xSZq+Il\ndBAxlBaJRIL44KDwmWhvh8yeuEm4gdD0gVpRUiREJiFh3B94w865c+Lc3ncwew3X/OijHLl2TZAk\nSeLC74epeJS3i0r5P56avyGOrutcunSJ119/nUAypWA2m7n//vvZvXs3pgW6J1dXi8P3+YR0O1/S\n4syZ2eOfJcHiSZIWeUq9Rj8ELY7NBj3hvUD23GqQFiUlC1ORrFqV/l7OnhVK17wIXOdKseDU4lz/\nuI1fvLMeXReD/eOPJxUpV7vEto76xQWGecJiETGjIRRaFjQ1CdJiYkIMqhktqFRV5Xf/8FvsnWX0\nT3gZcj1BmV3izBk/HR2FNXHyekVJod8vSkQWoKYuPBx1UPkADL4pHNsH3oC6Q/m91sjCKSUFzxhk\nKg06OtxEImI4u3jRT3//8ppqGWUiFy+qvPxyCxcuhCgrW5oTuaYJEiQUEguV+f4ePx4iGEwrGsLh\ndMBVVGSjokLN8hRa7oCsszOE0+lm3TpBjN4h+GzEI4uFPF1pkRtr1sDJzQ3ol2T6+jSqP2nDNgtp\nkUjAyy+nz6GSEkFYZHYDJJEQpAWIwDfTJDPlY5Fkzuq+DJZFKn8yg+q+vsUF92YXDoeD9evXcWJE\nJRqV+MUvfsG3vvUtLARFCQuAUpTyEbLb0358y6a06OyE3/yGlsuXeeXaNTCbsW7axNN/+ZfU56NA\nsHhEC9lClodoCRg+DiPvpQOyotVQ98TijKCnw2wWgdPk5IJJC0OFEIulb6qa/X8sBgwN4Tx1DEv3\ndXp7+xgaWkFMKeVi6f20WbYT7zWhaRCN7uX69RbWrt2VRRzcDgjbhCJFiYcxq+G8CQVJmkkUOFQ/\n9e3Hqeg/hyzpyF6QzDKTq5sJbtlHaVERlbOQDPncMkmK+RSvVVUOTpzwY7XOHA8ikQkaGm6DzP8c\nbU8XDU0TpUnvvJO+rpqahG/IzSqRgmyl06fM+GpW0kLX9W8n/95/047mVqAQSgtNw+rziVTs6Cgx\nVZ3pMyFJ6fINmw25vBy+8pU0OeFwzLmYMILeUEh0IMpcKCuKwsHnn8/yzzgWV5hY2cDOhw9gtc4d\ntPv9fl577TWuGRIHoLGxkcceewxvJiO3ABglIu+9J0hFv39+PxnD6wGEWrCqkEbYRu1nbGLOtqGA\nkFqOixoVS0kdU7qQUhgLdF1PkxYLVTUahpxvvimM0Pv7RRepeSEr6I4VdH/SQVtHO7o+ra5YnRLu\n7CBKB5YZHo8I4pYtSTJd4lBaSjwu3u83L52h6LKTaCjOgHUVSBI1NdkmToXyDJAkUebw8cfC9zNX\n956bivI9wtck0AXjLcKk1Z2jfGY6CtzuVNdFoKiq8NprLfT17cVkcqeuEa8XbLbl93AAOHhQ5cc/\nPoLfv5fhYTc7dghizXAi/973DqKqyrwEhHF/odNAIpE9ltjtDhTFj8fjRlFAUdLj+s0IyAwlh9e7\nfC1WbzY+M/HIYjGjPGSWzSTYd9DGJ4frcPt76DzWzoYvzdxOVUW3EaPqo7xcEBYzCPrLl9MShOlZ\nhoEjIlEAwsR6Ke2aa2vT7EFv7yJJC1E7Vl5Wzt5dG3nr1CWGh4f57cu/4cltU0iGSWjdE0IVgvi+\nnE4RcxVcaaGqcOwYfPghLf39grDweLBu2cIz3/52/spWY0yPFagTR3QMel+EULJvtWyGqgNQmm+G\nRSAXuZBJMDgDHkz9k4RiPm54Z243GxmhqnN7B1ojfhq63qZy8DyBJL/Z3a/zsf1hWt2fQ9WtkKVa\ncNPXFyq4hcBiF/+Zt6JeD+Vvi1O/av8EVNtnEAS5brKc8VMFg3DihKi7rkxAJeLJu+4S0qub2Uo1\niQMHmmltPcL4+N4s4iIa9VNaepIDBw7e9GOagUK3PfX5hLrC8A4xm4VE+t57b75ZmqKkScM7hbSQ\nJOlJcmc0ANB1/cVlOaKbjelGnORpoBmLidTrlStw7RrWjo7UyRiNx4XPhN0u6gnKysQJknFiauXl\nC3KOy1zADw3NDB4URWHXo48CYnF9fALKSDdiyAVN0zh9+jRvvfVWqqzF4XDw8MMPs2XLFqQlXkgG\naQGiRGS+LHVra9obqaAqC0i3PdU1UP1zd1AYbxHtRQG5Yg9ut8T4eJq0GBlJL2wWU4q5dSu89ZaY\neM+ezY+00DR4/3wT8b4O7CYfXuc4T37Dm35tsDu9sXPVwg9qgfB4xOk+F2mxmD7c4XCyE9NYOY6R\nImJjUwwNttPy3k6mpsR3duW9QR4ZFamQIe9KbLb0vLscJk719Sq//GULPl+ISESmsvIW9hOXZJGx\nbPuhOEdvvAL22hkZTCNgNII+s8+HrkIoXsJkPB38Gc9Pvz/fY5kNME6fDhEKZb//cv4e0/H++y1U\nV+8lGHSjqqLDgdkMsZibSGQv5861sGZNYUgTu13wy06n+OtwwNiYRigkiBJFAbu9mbNnjxAO7wXc\nyLKYRm9WQJap5CikMe2txGcmHlksMstDtLnTxI2NcHV9E3zYw0jrMPX9k7hq0gFFNCraZxsdrmpq\n4OmnZyl7Pn1a/HU4stuG+S/BmPFcHVQ+tPDPlAmLRWRuBgYW1UEEyGpXvnfXNrqGInR0dDDZfpgb\nTlmQBKU7s7t1IVSFU1MLU1rMO/fduCEWMGNjnOnv59W2NlizBmtDA888++yCSnFTsYymQiKU9Tnz\ngUFAx6I62thZ5MHDJOIqiQREpUrGXU8S7qogdm1+MmEh5ML6Kx6qBruJdvk4VYASfnM8woqe96jr\n+wA5eQ3oksxQ3XZes+/Erz2AYgKbKbssQZahqEimuTk/FUE+N5OpQOvQAQ8kicOqSh+sXwALHYnA\nqVPiltnVcP16QVYspaXgEqEoCs8/f3CWa+QWtzs1UKi2p7ou5GqvvZb+HSoqhNnmLfwN8HrvLNIC\nOIQIEiqAzwFvJR9/AHgfuDOChGlGnHMaaJ49y8EDB1Da2wVhkaEhsxjp1+Jionv2CJ+J8+ezfCYM\nTEQiOPJuHSGQeW4PDma4dudAZjvNVatybzMwMMArr7xCf39/6rGtW7dy8OBBHAUyZamqyi4RmY+0\nMAw4nU5hWFlQTG97OhtpoSVg9JS4by2D4vUzDJwz/SwWQ1oUFcHatSKLdfEiPPzw3L47Rtarr201\nO0qOYrPBs19sp6Q2gwkOdom/snlZ/SwMGCWpk5PiMjBPG0lm68Pd2+vnzJkj/P7vH2RqSsHnS7eL\n9vkys9wS68eaqBo8R2Kkk6mKBHpSAl0WHBZbSDBVuoq1a7MDhEJ6Bqiqyuuvp00Nu7vFPLbUfuJG\nzWw+ZMHM593YYoeoUX9JIhHG/+GLtIb/A6oqZ73GCBjNUoS9XkHItgc99C6+Ci4nNC37+zabs5WV\ny+3h0NkZorbWjd8vCMVQFkfixufLTZpIUpp4MEiIuf7a7blVNmbzdKmrwq5dB+noaGF4eJS6ulHK\nyztvWkDW0OCgr08cz/Tr8lOMz0Y8sljkqbSAZJL1iSbaP3wbTYPzL7az53vbAJGU/elPhQIQRPzw\n+78/S/fHoaH0ZLh9e3oSiPmgL+ljYbLBikX6WExHXZ0gLfr7Fyd5y/DykbUwX/nKV/i3/++/sErq\npL0DHJ56vJtnltIa5r75Ki1mm/v6+vxcuvg6z2+zYf7gA9A0Prpxg9eGhmDHDmweD8888wy1s2Qx\nUuTCNJJA95dgHxVfyWBwgpDuzFuxYPw1E2Sd6xXKLFdS79cb/hydoQfR8moyuHBEbCKIsMamkBMq\nslXBYiF1UxSy/p/1MSlO8dWPcJ19F3N5GFNVkjTYsB7zI/uRK8oY+/u3SHrK50R5ucahPCstbyqm\nG5bmA1UVwfSJE9mywYYG0SHlNvEvUBRlWRWYS4bJJOThPt/i256Gw6IzSGtr+rFdu4TC4lZPzl6v\nWCzeKaSFruvfBJAk6SiwUdf1geT/1cCPb8rR3QzIspiRo1EIh9MGmjabmCVCIRgbwz06yl6fj5bL\nl7MNNBUFmpqwbt8uTnCLhdjWrTRXV3OktzfLZwLAH41ysrSUgwfm95nIhMORVvMMDs69rUFaWK0z\nSyxisRjvvPMOp06dSvUq93q9PPbYYzQW2O4/s0Skr2/uEpH+/rQxuMF4FxSZbU+j4+Ca5bP6L4Ka\n9CEp3wOSlJo3DKWFEafZ7YvvkLV9uyAtYjExnm3blnu7cFh0CBHJpQqsThfbtgSwym1AhhzFUFo4\n6gVxscwwvhNdF79rVp0z6T7c8bib9nZxGYXDEIm4icX2cvXq/NnvcU8T9WPnKLbH2FnTh23dSrxe\nuHZlAK8ZNKud1Z+rMDrSpVBIz4CjR1sIBPZSUuImEBBzl9sNmuZmeHgv//iPLezcuStvtULmfW1J\nh7mRtc5mamwtQDf20HsMh/fl3NImp2udI9rs5SFmczowNILD6fdzPRaJaExOijneZBJB/s001VJV\nGUkSRG4iIYbyzOP0eGQee2wmCWGzFabbTC6pq8mksGLFBrZufY/nn/+Dm5o5mk16+2nGZyYeWSwW\nQFoA1N9bw2CZhZ6rHVz719/w7pQfRdFob3dQVtaMyaSwbh089dQcc7GRZZAkMWmDIP17fp1ui1n3\nROG6FdXXi/eMx0UQlFddZQYyFQjxAA63ha/tVjh/WozFvzwV5w82Rygqyr5WDc/PfJUWxtxns7lJ\nJMS8l0iAfSpK5YkOzr7zCbW1dZwZHuB3sQSh0m3Ig3buW/MsJ07UzEkw5ILLVMI9ya/40pSP4djC\nvhev0sZ612+wyOIDRhPFXA58mYl47sRa5vi/KJIh+Zj9mgf7YTFn7PtTP3JF2YKOG10X7SLffFME\nZzbErb4eDhzIyihlErnTcdt4KOSCzSYCTUOCOhcSCVHH+s47QhpkoKZGLJIbGm5Rz/ZPMUpL01m1\nhaKrS6ipDF9Dl0t4V9wunWqM8pdQSKhyCtL1YPmRz+qmHshcJg8By9yf6CbDZhORbiRCqL9fkAyh\nkFhNZtDrbpNJGGg6nbBunbg1NoKiYOntTRkyRKPRnD4TRpnJwcwykwWgqkqc/0NDc29nkBYrV2YH\n5W1tbbz66qtMJFffsiyzZ88e9u3bt6xGefmUiOSKfwoKpVgEdnpidpdtXYfh5MEqRVByF5BWFYRC\n4jTJ9LNY7BywZo1QXExNidMmF2nh98O//zupDEFDg8Q99zWhBM5DsFMEiLJJdDqJJDdaxlanmZie\nAMgkLRIJOHkyxKVLIvs9HWZzOvstSWlbF683+6/H0oj1b5IvWtMGD6wEXcfiGueG7sBfsn7GD1Do\nAKSzM4TV6k55eAQCmeaGbgYGQjeNqDaZssmCoOVhrLZuHKZRSsuOU2VpJG6pm0EwuLQJyoPiVGms\nLkF2zSQgFGXxC/jR0VtrqmWQIhaL6Dg0HeXlGvfcs5zvf3tJXacfzx2GOz8eWQwyieo8SAs1kaBN\nmyLsL8NrdnDszD4mAzLBoB+7/Qjf+c5BnnxSmX1MiETSA+GaNekJYfBY2guh7N6ld5nIxHQzzoWS\nFrI1HQPEgzD0FiW2GGvXrOWNc1EGwyZ++ctf8s1vfjPLeDxTaZGPZURnZwhZdtPZKQ5Ti2tsmjzF\n5om3SOgJrisqJ/xhDutFqBYX5kE7d9/9DH19i1NIhrX0ZJxJUGcSBrnIA4tZpVo6hlf/MEU4xxyb\niHgfo8lmz/napcwTM6B60k1u/D5YCGnR0QFHjwrljYHSUrE4Xz8zLvhUeCjMBo9nbtLCIG/efjt7\ncV1eDg8+mPP7+B/IE4tpe5pIiN/i5Mm03HXdOuGYfyt6ss+G6Z4dhTZ1WSbkQ1ocA96QJOlniJzm\n11i27r23CHa7WCGGw8hGsfbVq9l6QIcDysqQGxrgL/9yxshtzVBTGP4QmT4ThUBlpej6NTYmmPdc\nsfDERFoRYJSGBAIB3njjDS5evJjarq6ujkOHDlG5zDVVVVXi2hgfn71EJBwWZRIgxtf5DDsXBUkW\nGZ/o2OwdRCaviq4MAGW7U4Fg5gK9pyf9/S6ltbgsC2+LEyeEimJkJFu1MTwsCAuDpN28WZC05qkm\nCJwXZqHhPkFSBLrSL7wJfhYALpfK9evC52FyUqapSaOqyoHb3cz58wotAJoeGQAAIABJREFULdl9\nuA15vZE4qKiQ+f73xW89u8rXKZwEBwaEGedDD8H4OGur3Yx3t3PNsT9r6+UIQIzShvLy3KXUmaUR\nRtuufBUK8z0/nVSY+T1ZIPwktP0T6AnusrwAa/5YtEfNxIgPjNhudUl+o/4CcKsDwtshi3a7SV0z\nj+d737vFB1NY3PnxyGKQpbSYv/XB0aMtjLofocL5NuFwGL1/gKi1FrPZjcezF6ezBVme43y+cCGd\n+jcMqCavpksrHTXCvLGQ8HjERBIMisE4737hAmo8Tsv5bkIT/cjKJ2jxMA5XCc079+KNVNH18Xl6\ne3t54403eDQjbjOUFomE4GrsczRwCIehtVXm6lWxvUv1cd/oS1RGRaZDR+Kw2U2v2QqSjNls5557\nnqW0tHrBioX041Y8A3bMhLm31Idcnwe5EB4UZpsRUWqJyQI1j0LJ3TdvgbuY0oehIUFWtLWlH3M6\nhUfDtm2zBhO3G7G8IHg8Qoo8/TvSdbh+Pe3qbqCkBO6/XzD4BWOYPqPIbHsaDM7fanl0FF54IU2m\nKYqo/25uvv2IozuYtPhT4MvAfcn/f6Tr+kvLd0i3AMYsFImgKYogMIwUcUWFWP0nfR608vKcA0Em\naRGNRpflMI1SD10Xi9pciYZMP4sVK3TOnv2Yo0ePEk7WtlmtVvbv309zczPyTRjQjBKREydE1mFi\nIrtFNwhFm2EPUnADzkxYPEnSIscEqeuivReIOlxvWu6RebyZLQSXQlqAmGNPnBD3z54VYxuILhU/\n/3m6+cyuXeI5SSLbJGyqXZAWRmmIbBamZ8sMVVX5l385Qk+P6BjR0yN+24EBPzbbEXbtOogsayiK\n4BxqamYqz8rLNfJqTNPUJCaAgQEhdenuxmQysWtXE+F1UXT/W8sagBhZ/OJicW5GoyIukmXxt7JS\n40/+RASQWY7dNwv2aqjaL9qfxnzQ/xrU/172Nka7U5MFTIVfwN/qgPBWkyb/AzcVd348shhkkhba\n/EqLzs4Qgap7WFvyNuEw1ITbGbXWsmoVrFw5j3murqelkR6PkDvHJqDvN+IxkxVWPFX4MkVJEtL/\nK1cWbMapqipHfvUD9pb04a6MAOLz+UNTHPloiANPfZPBoRH6+/s5ffo0dXV1bEnKtjKTo4FAbtJi\nclLl5/9wlI73O/Fdb8WpfoKLBPtNk1QUJTCVQNhRyuvF9QyGXuJzO7y4XHa++c1nqakpQIufmAdC\nYTBNpNULuaDrglgafDOtyHHWC3Nnaz4TcgHhconaI6Mt2Fzw+0X2+vz5dPbaYoHPfU5kwnKarmTj\ndiOW84GqqrRcu0bo9GlkTUOTJBxNTTSvXYvy7rvZBmtOJ+zbt0z11Z9RZAapY2OzkxaGq/7hw2mX\n8upqYbZZtsCyp5uF6aTFpwTzntm6ruuSJJ0FpnRdPypJkkOSpCJd16fme+2nBsaKKhzG0dCA/8gR\n3CAmyaam1IA4l4GmJcNJcblIi+lmnLlIi+7k+jUeH+XIkVfp6elKPbdhwwa+8IUvUDyjb9nywiAt\n/n/27jw6rurK+/53l1QaypLKlpBng2WMsc1oBMSATBwHqcGAQ9IhQ7+dQNIE0p0QnHQGknSARcZO\np7ud4UlW6NB5kqzMoZuHMUgONpjRILADtgHbyMbzJFtY1uAr1Xn/uFVSSdZkqSaVfp+1tFx1b9Wt\no6Prql37nrMP+FNELr20e59z8OKL/u1TTvGn3SVNrK5FX0uDtbwFLdGiGmUX9bhaHZ+0eC1apyo3\nd+TLCZaW+itTPP54PevWtfDqqwEOH46wfXuImTP9ecXV1X5/dTU1d5z/RbV1DzRvAZZ0F+EsnJaS\nehZ1dfUcPlxFUVGY1lY/uQyQkxOmtbWKI0fqWbo0xI4dTRQWjvDq9+mn+/OLnPOHhEazcjlFRVzx\n/y3liiRnCeKv4o8b1zOAbWs7wrx5oQGvvKXEKQv9ZVCPboHDf4Wi2TAhbp5EbLnT4PikZVXSGRCm\nO2kiqTMm4pHhOMmaFp4XoD2/hI4J5RQ3H2BG+xZaz7i8K6YYsHju9u3dcxYvvNBfkeutP3WtuMX0\n9wy8OtdITJ/ufwg3NflzK4uLh/S0+mfqqJrZSLgjBO3dQwDDk+ZQVd7O+rWr+OAHP8hPf/pTWlpa\nePDBB5k4cSKTJ0/u8T3l2LGeIyJbW+Gppzwe/Y8VLDl6iDNzC3g70EzZ0ccoCBzmqcJCTg0tZN+M\nS1gdHMemLY9y2mlBxo8P8dGPfpTJiVrXPTge2N39Xt8X723YcT80v+nftwBMfCdMXDTwEvDJYuYn\nvQ4c6D9p0dbmB4/PP999ZSsQ8IuCLV48+JXvUaxrUYDNmwnHqku/9hpNtbXUtrVRs3AhwZwc//vL\npZf6V7cGquguJy9+3nN/y54eOwYPPtj95cDM/3ssWXLyxYJTyDOjfudOWnbvJrBjB5FXXjlxlcwM\nNOg3HDO7GfgEUAqcDkwHfgKMcA2rDBL71tHaSuWCBdS2tlLV0UF4xoyuhMVgBTT7mh6SaKWl/mgj\nz+u/GOfWrR1s2/Y0zc1PkpvrBy8lJSUsXbqUucNZ2zwBJk3y/+8fOuRPEYlPWmzd2p3ku+iiJF+p\njgVSne0nLg0Wq2URyIWynsNOY4UF45d6nDZt5Mlsz/PYsKF7ZYrnnvM/vzs6mti3r5ZvfauGCy/s\n482j+HQ/adG6B9oOdg/xLJo5sgYNUazOQyjUXZzazB+UNG1amFmzWrjllkWsWJGAq98zZnR3/tat\n3UOJRlJQ5CSMiqv4Zn7Bu80/8edq737YH3ETu3IWG2mRqIJ4GWg0XkWTkzcm4pHhiF/ydAjTQ2Ij\nyA6Xns70lgNMYCftk9rooKDH/r54zzxD/e7NtFgTgR0hIpt/S8gOU3n2LIKTF0J4/sh+l4HMmNF9\ne8cOmD+012ppbCB8ej68HfelLn8ChKYRxmjZ2kA4vNRfUeRXv8LzPH7/+9/zsY99jOee28jatS1E\nIgE8L8Jll4VYtKiS+vogzz0HO1+p42+ONlKUk89E9nNhyS4aW/fT0RHmctfJ7yfMZH9+mDfeeJDC\nwrc4++xzuOGGGxI7NTf23n78SN9z75s2+stjxxJL+aX+iLwUjMwcUH9Ji44OfzTPk0/2XAFj3jx/\nmmimXr1OoK5FAeLnS2/YQBio6uigfvt2Fn70o3DZZQPPWZLhGx+90ONc36MRtm6F//3f7iq9JSXw\n3vcm+erryHUlxHbv9hNiBw/CgQP+KpkbNlCzfHnGJi6G8rXrU8DFwHMAzrk3zGxiUluVanHTQ4Iv\nvEDNO95B/Ztv0nLuuQTy84dUQDMQCBAMBvE8L2kjLQIB/4vhrl19F+N85ZW3WLnyQVpaDjBrFpgZ\nF198MUuWLOmRVEm12BSRJ5/02x4/RSS2zHswCOedl+SG9F72NJa0aN0HRzf7tycsgGDP7H2sWGT8\nkll9JVxPVl1dPcGgX2W8o6P7+Pn5YWbNqmL//nqgjy9iRaf7SRbnYP/q7u0pKsIZuwp3ml8Xk5IS\nfwpILMnf0RFI3NXv3Fx/etbmzf4wndj/rf7W8k2wUXMVP1jkJy62/dpPyu34H5j1Mf8KWmw6VLKu\nfoqkTvbHI8NxkiMtYiPIGktnM33ncwSIMP7INg6eMnfAkXBeYyO1df+XqnkthGdOh4odcORVmlo6\nqH2qlZpPfoWkviNOneoHQpGIPydxiEmLgItebYgtexrIhZK5xJaeiu2fNWsW7373u1m5ciUHDx7k\nllvuZvLkf6GlxQ9Ydu2CX/+6iR/+4DEWnX0R4ZaDTNv8KGd7O5lY2Ewo6GH5EJo+gb3k0RiazJ6m\nP3PAe5LTTgty9tnn8PGPfzzxtcRi7+2uEzqO+oXHwf8s2P0oHF7X/djSBTDlyhNrH6WY53nUb9ni\nT30wI9LZSWjWLConTSK4Zk138TDwk1U1NT2TVlmupaHBXxSg9zJjZoRPPZWW887zC49K8uTk+MF/\n7xVEOjpg5Up47rnubWedBddcMyoSSF0JsZKS7qX9gHB+PlWNjdTX1SW0HmMiDSVp0e6ca7do5tbM\ncvHXS88esekhHR2wfj3BnBwWXnstfOhDJ3WYvLw8PM9L2kgL8OtaxJIWsYR6W1sbK1eu5KGHXiQ2\niuyMMybxkY9cy/QMWZN5/nw/aQHdU0QOH/a/i4JfMyjpK+7Ez9lsb+y+ynDwaf9fMyi/9ISneZ7H\n9u31vPqqf7UlEIgwY0YIz6sc0ZfWhoYWCgvDTJrk/03BT96cey4UFw8wrzg0AwJBiHhw5NVo23P8\n7SkQX+ehr9UaYvsTdfXbmzmT+kceoeXwYX9eZyBAaPZsKi+8MCVJg1FzFb/kDH+qyMHn/KlO+5/w\n73dG34+C2TvSQsaM7I9HhuMkkxaxEWT7Ot9BJJBLINJBaeMWdhVPGXAEWf2v76FqbgvhwlyYcgq8\n7Q+JDhflU7XgVOqfW8XCdyYx2A0G/SBo9+6TqmsRsejnROFUP2ERDPurifTeD1x22WXs2rWLBx6o\nY8eOU8gPvsR47xxK2/dQ9tIeyo7vYXxbA+W77mfmzOnsydvMzMK27sENwSCBOXOYWl7O1u3bsT0H\nufjisxg3bhw33HADEycmNsfmeR71a9fTsnUtASJENq4gNOVCKs+bS3Dvg91J69xCmLYMwvMS+vrD\n0XWld8uW7qkPGzfS9NBD1HZ2dk99OOUU/4v5mWdmXiHDJOtaFKCgwJ8G09zsD1ueORMKCwlk8NSD\nbOF5HvU7dtCycSOBDRuI7N9PaMIEKhsbCcbmReflwdKl/lXXUXKOdiXEYgmW48f9ysE5OYTz82lp\naEhvAwcwlKTFE2b2VSBkZtXAPwEPJrdZqeN5HvUvvdRd6CYQ8E/KG2446SsG+fn5HDt2LGkjLaC7\nGGd7Oxw+7NizZyOPPvoozc3NHDkCgUAus2cv5vOfv4RgMHPe1PqaIlJf311T6eKLU9CI3iMtwB9O\nGfviHz7rhKvRnuexYkUtDQ1VtLR0D9N77bUmVqyoZfny4V9tj41YmD7dL6yal+evEhJ7H+l3XnEg\nFy9/BvXPPEbL0cN+oJJ/CqHGOiovTf58tFSu1uB5HrV1dVRt3044Nh8nN5emV1+lds+ejB7GlhaT\nr/CXw23dh7d7NfVrX6blrWgwW9pBaNrhlJwjIkmS1fHIsMXXMooMZXpI9wiyyN5mQvv3MuP4AcZX\nlVFd089nWmcnLa8/T/jMXCgOQedbfuIcoHgO4YLxtGxNQbA7Y4aftNizx7/QNIR5mqHSCpqadxIu\nyoeCnnUkjhxtI1QaHc7d2YkdOMB1M2ey6q39TD+wH7frBS7rnE95oHt+e16eUVrqccEFsOrFHIxx\nfn2NoiJ/OGxeHk9u387jDQ24UCipCYvaP66gatpOwmdGv/yX7KLp7depvfcANYvf4ceBRbNgxnXd\nIzDSrMeV3piNG7unPuzZw8JPftKvXTFGV8CIxP4Pmvn90NnZY8nAiD7Dk6orsbZ9u59Ya2+Hl1+m\n6Y03qM3P9xNrp50G73sfQ6sqnzm6EmLxo0JaW7tqxHTtz0BDSVp8CbgJeAW4BXgE+FkyG5UqXSfl\na691Z3uBJjNq//Snk/5CFCvGmcykRWxUYVtbE/fe+zDHjr3RtS8nZxYXXXQN555b2udyqOnUe4rI\ngQN+sV3wyxMkeeVVXyAIwWLwjnYve3rgGb+QGED5ZSc8pa6unsbGKkpKwl3TN4qKYNy4MI2NVdTV\n1Q/7KnxsREJhYXedj/hEbX/zij3Po3bVC1RN3E54avS/8LgATayh9o8bqLk+uV/kU1nnob6ujqq2\nNsLjxnVPDQmHCRcUZPwwtrQI5MKM9+Nt+jG1jz9L1Rmt/pcMgLJWmtpSc46IJEnWxiMjcpIjLSBu\nBNn4CNTW+hsvObPvtdQBXn+dwPEWyDkG41vBiwa8hVOgwP8A75qGkUzTp3cXZty7178/iMpLq6n9\n4wa/GGdRdIRFpJOmfYd5en0ONWe0wz33+ENYOzvJB94TDHHf8QN0uE7WtW/ikrxKxoeKyJk6icik\nKeyZOgW7+RpClRfQ9Oyz/pXLqCe2bWPVtm20dXRQMmUKN954I+XxFTwTpKvA6LgSiJa24uhmwkSo\nmt1B/avbWLj0n/wRdxl0FbjrSm/vL0c5OYRnzqTlnHP8Iq9jWKiigqadO/1+CgR6JG8GWhRAEqNH\nTZEDB/yk0ZYthAMBqtraqB83joUf+1hGF9vsT1fCa/x4f7h0YWGPoe6ZnBAbMGkRHXr5qnNuLnBP\napqUOl0nZfySAEB41qxhfSGK1Y1I5vSQ8vIIO3eupaHhcWbMOB5djTXEZZddiXPnYGapmu5/0ubM\n8bj33noOH25hw4YAx45FmDAhxLJllZDcmbDd8iZEkxaH/aKFh1/2txfP9oOvXmJFJ+OnrsTqIuXn\nD7I03CDiRyz0jicGGrFQ/0wdVXNzCLfH/ffNG084L5+qmY3UP1OX1CG6qazz0NLQQLigwM9kx9a+\njhZEyfRhbGlTUE79zvF+wiIUd47kFBIuyk3JOSKSaNkej4zIMJIWXU6PW0Z7y5b+hz2+8CSRvH1Q\n8Hb31MqCU/zPzqj4aRbJ4k2eTP3mzf50wcOHicyZM2jV+6Bz1LzjfdSvfICWPZsIHG0k8nYrIRtP\nzaRZBNevP+E5E8aN47yZ81l18HXay4qoLSrlvMtuIyfPDwbKyx+HadOonDiR2tdf92PJ/HxWb9vG\n6mjC4sD48XzjG99ISsIC4gqMgr+kdefxrosw4ZIwLe3zofySpLz2SHRdyQ2F6KroPWWKP/UhL4+A\n04yvyupqajds6DqvYgZbFEAS44QpFDEFBYTnzaOluHhUJiygV0Ks1yiRTE+IDZi0cM51mNnrZnaa\nc257qhqVKl0nZXySYdw4KC0lbHbSX4hiSYtkjbTYs2cPDz74IDt37qaz05/idv7551NTU8PmzaGu\nL76JKBKZaJ7n8Zvf1LJ/fxWeF+6qvdHa2kRdXS3nnJOiooZ5pXDsLT9pcWht9/DWPkZZ+O32s9uh\nuPxB/BKoAy4NN4jhjljwA5UwHMz3C22Z+XN08ecWp2KIbqrqPHQFN30kLXrslx5avBzCp0yG9ui8\ny0Bu1woDqTpHRBIp2+ORERlJ0mLiRH9qw9GjfjX83kkL5+DNv8CRPxLK76QpUEg4Jx+Kz4CCcmLF\nLHtMs0gSz/Oo/fnP/ar3zvnTRCZM6Fn1PjYCY8+e7p9Dhwg6Fy1rPRWKp0L8aqn5+f6X5rifxuc3\ns/vpsxi383kadzwDtPLG1keZO/c62tubui4qBINBapYv58XaWp6preWVvXtxoRDhqVP5xte/zpSR\nro0+gB4jW3JC3fWLQtOhaBaBw5lZFLDrSm4g4I+oiER6TPPJ5Cu9qRI7r+rr6mhpaCDgeUNaFEAS\noyu2LC7uLv47aRKccQbk5o7q2HM0J8SGMj2kFNhgZmuBY9Ftzjm3LHnNSo0ehW5iy9rELaV4sidl\nsqaHHD9+nNWrV/Pcc88RiUQoKgKzUubOvYbrrpsFdK8EmZfnf+Zmmrq6eg4frmLq1DDb48LNU08N\nc+TIyKZZnJRYzQrvKBx83r8dmgbjZvb58NgUjXHjYNYsfzRj/GpbAy0NN5jhjljwAxWDgolwbIf/\nO8UFrSkZopsiXcHLKaf4V2Fyc/0Pkd77pYcAHVByJjS+7QeyOT1H7WTTOSJjStbGIyNiAf/HRYa0\n5GnP55o/2mLdOmho6CrIBvgFq3c9AOvqINBJ5YwJ1DZOoWr2DMIF3atsNTW38/T2MmquT26wW19X\nR9Xhw4TLyvxl+t5+GxobCR89StWmTdRv2cLCwQKgwkJ/FZL4JMWECSdMn6j+m/Fs2FSLc5dy9Ohu\njhzZxr596ykoCHP22a09Lirk5ubSFgpxdMIEZi5cSHFxMTfeeCNlZWW9Xz2heoxsKTodWnf5U3Wi\ncU4qRr4Mh6Y+DE0wGNT01zTpii3z8qCy0n9fjKvBMppjz9GcEBtK0uJfov/Gv6NnxditrpMutmSD\n50HcML6TPSmTMT1k8+bNPPzwwxyJLv8UCAS45JLLOHLkcnJygrS1+TmXWCLg1FMzc8RSbJrFxInd\nbTXzY4eRTrMYKs/zqK/fQMub0cKEBAgVT6Dy6vcR7Ge+Z/wUjlNP7bkvEUUnhzNioSsQGVfhjxwJ\nFve9Pwv0CG56zXtScNO/iAX9Gi7jz4OWHSdMfcqmc0TGlKyNR0bMcqJJi5McaQHdSYvjx/1VOU47\n1V+FaN8q8Npg317ozCdYdgU1H1pO/TN1tGxtIOA8IhYkVFpBzfXJD3a7RseWlPhJi/Z2+OtfAQgD\nLceP97xqU1x8wggKSkqGVN8h/qLCxInlPPnkWo4fP0Zn5y7+9m+/3PW7OudYtWoVT0aXRyspKeGG\nG25IesICehUYDRZDcG7XvlSMfBmu0XylV8aGHrFnrxIC2RB7jtaEWL9JCzMrBD4JzAb+Cvy3c9l1\nea7HSTmh56oRwzkpEznSorm5mccee4xXXnmla9v06dNZtmwZjY0T+d3v/G379vmj5WNLCGdqPYv4\naRax1ZtOOcUflRm/P3mvH6uyva27yjbQ1OZR+9ifqfnAvD4DrlQWnRyqHoFKr9VOMjlQGQ4FN8PT\nfY6Mg5K5PfZl2zki2W8sxCMjZjmAN/ykRWy06eYXoaMOWqLrcO/bB8cmQcssuGqZH+ymqR5O1+jX\nXvEaAAUFBE45BZYs8ZMTkyf3GJU3HLGLCkuXwgc+cB4///nPOX78ON/61reYN28egUCAN998k8bG\nRmbNmsWECRO48cYbKU3RagJ9FhgldSNfhms0X+mVsUGxZ2YaaKTFL4DjwBpgKTAfuC0VjUqVRJ+U\nsZEWnucRiUQIDGOpJuccL7/8MnV1dbS2tnYd94orruDCCy/EzIjmRgA/nmhq6r6fqUmL2DSK2Coi\nhw51L98avz9Zuqpsh8JwoHt7eOIsqsoO91uYMJVFJ4dqtAYqw6HgZnjG0jkiY0LWxyMjFsiFToa0\n5OkJQiGYMhEOPw87X4JJF/jb80thWy4c6/ArUJ9xRkKbfLK6Rr8WF/vrg7e0+FdBioshGCRSXg6X\nX56U154+fTrV1dV87Wtfo7W1lQMHDlBSUsKOHTvo6Ojg7bff5mc/+1nKEhYQ/Xy8Pn0jX0ZitF7p\nlbFBsWdmGihpMc85dw6Amd0LvJCaJqVOok/K/LjEx/HjxymIX3JiCA4ePMhDDz3EtliBCmDevHlc\nddVVlMTNpQqH/SkhbW1+vamYTK1nAT2nWRQW9lylLBHTLAbTXWXbQSAHIp1+te2CSYQtMGBhwlQV\nnRyq0RyoDIeCm5M31s4RyXpZH4+MWKyu0XBGWrTshCmboH0bNOMvJzp1MbRVwN5f+Y+58MIetQfS\nocfo2PjiUqRmyHZjYyMlJSV4nkdTUxNN0StG48aNY/bs2dTX17M0xZ9V6Rz5IpLNFHtmnoGSFl3p\n+mjV7oS+sJmVAr8HTgO2AR9wzh3p9ZgZwC+BifjzVu9xzv0gke1I5EmZFzcEor29fchJi46ODp5+\n+mmefPJJOjv9gKOkpISlS5cyd+7cEx5v5hex3b7dH2kRHZCRsfUsIP3TLLoLDxoUTPELVhWd7hcv\nY/QVJlSgIoPROSJZJGnxSKbEIiM2nKRF53G/bsWh52B8NCHRUQS8Cya/G+67z9+WkwMLFiS0ucOR\n7iHbDQ0NzJ8/n5dffpnm5mYACgoKOO+88ygsLKRBS3CLiCTNQEmLc83saNz9wrj7zjlX0teTTsLt\nQJ1z7rtm9qXo/dt7PcYDPuucW2dmRUC9mdU55zaN8LWTovdIi6HYvn07Dz74IAcP+ksTmhkXX3wx\nS5Ys6XG83mJJiz17/JV4IHOnhkD6p1n0KDxYPBuKZnUlLE7YLyIimSSZ8Uh2xCLRJY2HvHpIcwPs\nfMBf/hsgPAGOnwFHp8D2t2FOM2zc6O+bP9+fhpFm6R6y7XkeOTk5nHXWWWzYsIFAIMD8+fO7LlB5\no3gZRBGRTNdv0sI5l+xr9suAd0Zv/wJYTa9AwTm3F9gbvd1sZpuAqUDmBApxeo+0GEhraysrV66k\nvr6+a9ukSZO49tprmR4/d6IfsXoQkbhSEJmctID0TrPoUbwSeiQsVJhQRCRzJTkeyY5YZKgjLTrb\nYE8tNL7UvW3cDJi2DDashNdfh61boazMX+YP4KKLktPmYUjnkO1YUqSwsJDKykp6j/jRtDsRkeQZ\nypKnyTLJObcvensfMGmgB5vZTGAB8HxymzV8Qxlp4Zxj48aNPProo13DC3Nzc1m8eDGXXHIJOUOc\n31FW5rF5cz2HD7cQiQTIy4uwbl2ISZMq9cHZBxUmFBGRPmRHLDKUpEXTa7D7YfCig1QCQZh8BZRd\nBBbAmzmT+oceouXwYQKrVhFxjlBFBZWTJ6OoAioqKti5cyf5+fknJCza2tqoGOXLIIqIZLKkJi3M\nrA6Y3Meur8bfcc45M+t3rfXocMw/Abc555r7esxdd93VdXvx4sUsXrx4GC0emfikRV8jLY4cOcIj\njzzCG2+80bXt9NNP5+qrrz6pitOe5/Hb39ayfXsVubl+fYjCQnj66SY2bapl+fL0rGqRyVSYUETG\ngtWrV7N69ep0NyOjpDIWgTTFI4FoONfX6iEdx2D3o3Dk1e5txafDtGu6ls32PI/a2lqqtm8nnNsd\nGja1tFD7/e9Ts3z5mP+crK6uZsOGDTQ2Np4Q75WVlVGtZRBFRLokOh4x5/r9fE4qM3sNWOyc22tm\nU4BVzrkTqk6aWRB4CHjUObein2O5dP0e8RobG/nBD/zaXNdddx3nn38+AJFIhOeff55Vq1Z1jcAI\nhUJceeWVnHPOOSdk7AfzyCPPsWbNPNavD9PS4m+rqIDTTvMLWy5atCmjVrsQEZH0MDOcc4mtpJ1F\nEhmLRB+Xnnik4VdwdCuEpsPsm/xtzsGRV2DPo9ARrdidUwBTr4TKJ1OUAAAgAElEQVTx5/lVvaOe\ne+QR5q1ZQ/jll/2lycAvwHnppTR1dLBp0SJV0sdP7tTV1dHQ0IDneQSDQSoqKqjWMogiIgMaaTyS\nzukhDwA3AP8a/ff+3g8w/9v8vcDGgYKETOB5HqtXr2bt2rVEIhGOHTvG7t27Ofvss/nzn//M7t27\nux57/vnnU1NTQyg0vGU+GxpayM8PU1REV9Ji/Hj/3/z8MA0NLSP9dURERMaCUR+LeJ5H/Quv0LL/\nVQLBDUReaSdUMpHKGS0E2+JWtAjPg6lLIVh8wjFaGhr8FTlKSyEWr0yeDDk5hHNyaNHKGECsNpeS\nNyIiqZbOpMV3gD+Y2T8QXWYMwMymAv/lnLsauAz4e+CvZvZy9Hlfds79OQ3t7ZfneaxYsYL9+/fT\nEs0i7N+/n1/+8pc0Njbyjne8g5ycHEpLS7n22mtHPO/R8/wikuEw7N8PwSAUF5+4X0RERAY0qmMR\nz/Oo/eMKqia8QXhCix/VhdbRtPcNah/Io+ZdCwmGwn6yIjy/3+MEYitflJf7SQszmDbtxP0iIiJp\nkLakhXOuEbiij+27gaujt58CMv4beF1dHY2NjYRCodjQF7Zv3w5AR0cHDQ0NfPzjH2fRokUJGT4Y\nDPpLhkyZ4o/eLCqCQODE/SIiItK/0R6L1D9T5xeY7syDNqCjBd7eTDhkVJ3RSn0DLLzuU5BbOOBx\nIrHYZMIEOOcc/2pI3GjQiKY+iIhIGmXkh/Bo09DQ0FVNuvfqH6WlpZx11lksWbIkYfMdKypCtLc3\nEQj4ozfjl09vaztCRcXwpp2IiIjI6NHS2OCviGW9wrmcfMLTLqAlZ/KgCQuAUEUFTbEC4mVlUFLS\nte9IWxshrYwhIiJppKRFAnhxwyYLCgoAyMnJ4YwzzmDBggUUFg4eMJyM6upKSkufor29qcf29vYm\nysqeprq6MqGvJyIiIpkn4KLxR050NQsDQtP8ZUzzSrv3D6KyupqnSku7ExdRTe3tPF1WRqVWxhAR\nkTRKZ02LrBE/gmLevHk0NjYyceLEriWxEl1ROhgMsnx5DXV19TQ0tOB5AYLBCBUVIaqrtdypiIjI\nWBCx6Od9aAZYEIIl/k/v/YMIBoPULF9OfV0dLQ0NBDyPSDBIqKKCGq2MISIiaaakRQJUVFSwc+dO\n8vPzGTduHOPGjeva19bWNuLCm33xK1hrWVMREZGxKlRaQVPzTn+KSGh6j31HjrYRKh16/BEMBrWs\nqYiIZCRND0mA6upqSktLae81rLK9vZ2ysjKqNaxSREREEqzy0mqe2lZKU3OvaR3N7Ty9vYzKSxV/\niIjI6GfOuXS3YcTMzKX79/A8j7q6OhoaGvA8j2AwSEVFBdUaVikiImkSXdHK0t2OsSId8YjnedQ/\nU0dLYwMB5xGxIKHSCiovVfwhIiKZYaTxiJIWIiIiWUpJi9RSPCIiInKikcYjmh4iIiIiIiIiIhlJ\nSQsRERERERERyUhKWoiIiIiIiIhIRlLSQkREREREREQykpIWIiIiIiIiIpKRlLQQERERERERkYyk\npIWIiIiIiIiIZCQlLUREREREREQkIylpISIiIiIiIiIZSUkLEREREREREclISlqIiIiIiIiISEZS\n0kJEREREREREMpKSFiIiIiIiIiKSkZS0EBEREREREZGMpKSFiIiIiIiIiGQkJS1EREREREREJCMp\naSEiIiIiIiIiGUlJCxERERERERHJSEpaiIiIiIiIiEhGUtJCRERERERERDKSkhYiIiIiIiIikpHS\nkrQws1IzqzOzN8ys1szGD/DYHDN72cweTGUbRUREJLspHhEREcl86RppcTtQ55ybA/wler8/twEb\nAZeKhomIiMiYoXhEREQkw6UrabEM+EX09i+A6/p6kJlNB5YCPwMsNU0TERGRMULxiIiISIZLV9Ji\nknNuX/T2PmBSP4/7T+ALQCQlrRIREZGxRPGIiIhIhstN1oHNrA6Y3Meur8bfcc45MzthqKWZXQPs\nd869bGaLB3u9u+66q+v24sWLWbx40KeIiIhkldWrV7N69ep0NyOjKB4RERFJrUTHI+Zc6qdmmtlr\nwGLn3F4zmwKscs7N7fWYbwEfATqAAqAEuM8599E+jufS8XuIiIhkMjPDOafpDP1QPCIiIpJ8I41H\n0jU95AHghujtG4D7ez/AOfcV59wM51wF8CHg8b4CBBEREZFhUjwiIiKS4dKVtPgOUG1mbwBLovcx\ns6lm9nA/z9GlCxEREUkkxSMiIiIZLi3TQxJNwzFFREROpOkhqaV4RERE5ESjdXqIiIiIiIiIiMiA\nlLQQERERERERkYykpIWIiIiIiIiIZCQlLUREREREREQkIylpISIiIiIiIiIZSUkLEREREREREclI\nSlqIiIiIiIiISEZS0kJEREREREREMpKSFiIiIiIiIiKSkZS0EBEREREREZGMpKSFiIiIiIiIiGQk\nJS1EREREREREJCMpaSEiIiIiIiIiGUlJCxERERERERHJSEpaiIiIiIiIiEhGUtJCRERERERERDKS\nkhYiIiIiIiIikpGUtBARERERERGRjKSkhYiIiIiIiIhkJCUtRERERERERCQjKWkhIiIiIiIiIhlJ\nSQsRERERERERyUhKWoiIiIiIiIhIRlLSQkREREREREQykpIWIiIiIiIiIpKRlLQQERERERERkYyk\npIWIiIiIiIiIZCQlLUREREREREQkI6UlaWFmpWZWZ2ZvmFmtmY3v53HjzexPZrbJzDaa2cJUt1VE\nRESyk+IRERGRzJeukRa3A3XOuTnAX6L3+/J94BHn3DzgXGBTitonIiIi2U/xiIiISIYz51zqX9Ts\nNeCdzrl9ZjYZWO2cm9vrMWHgZefcrCEcz6Xj9xAREclkZoZzztLdjkyleERERCT5RhqP5CayMSdh\nknNuX/T2PmBSH4+pAA6Y2c+B84B64DbnXMtQX2TXrl0sW7aMdevWEYlERtxoERFJDDOjvLycm266\niTvuuIP8/Px0N0nGJsUjIiJjmOKR0SFpSQszqwMm97Hrq/F3nHPOzPq6LJELXAB82jn3gpmtwB+2\neUdfr3fXXXd13V68eDGLFy9m2bJlvP/97+fZZ58lLy9vmL+JiIgkWkdHB2+99RbLly+nurqaJ598\nMt1NygqrV69m9erV6W5GRlE8IiIi/YmPR2pqanjiiSfS3aSskOh4JJ3TQxY75/aa2RRgVR/DMScD\nzzrnKqL3q4DbnXPX9HG8Podj5uTk0NraqgBBRCRDtba2UlxczEsvvcS5556b7uZkHU0PGZjiERER\nge54ZMuWLcycOTPdzck6I41H0lWI8wHghujtG4D7ez/AObcX2GFmc6KbrgA2nMyLRCIRBQgiIhms\nsLCQzs5OVq1aRUvLkEfbiySK4hEREemKRx555BFUmyjzpCtp8R2g2szeAJZE72NmU83s4bjH3Qr8\n2szW41fr/lbKWyoiIklnZhw7dizdzZCxR/GIiIh08TyP48ePp7sZ0ktaCnE65xrxr1T03r4buDru\n/nrgohQ2TURERMYIxSMiItKbRlpknnSNtJAEuvHGG/na176W7maMOer39FHfp4f6XUQGoveI9FC/\np4f6PX3U92NPupY8TRvP86irq6ehoQXPCxAMRqioCFFdXUkwGEzJMYqKijDz65AcO3aMgoICcnJy\nALjnnnv48Ic/fFK/k5l1HW8wf/jDH1ixYgXr16/n4osvZtWqVSf1WsPleR71dXW0NDQQ8DwiwSCh\nigoqq6tPqt9Hcox09vvnP/95HnjgAfbu3cu0adP4yle+wkc+8pGTer3h8s/XOhoaGvA8j2AwSEVF\nBdVD7LeRPh/S2/df/OIX+d3vfkdTUxMTJkzglltu4ctf/vJJvd5weJ5H/TN1tDQ2EHAeEQsSKq2g\n8tKTPOdHcIx09ntMY2MjZ555JnPnzmXNmjUn9VyRbKZ4RPEIjJ14JBGxxEiPMRZjEVA8EqN4ZHQb\nU0kLz/NYsaKWQ4eqKCgId23fubOJDRtqWb68ZtD/eIk4RnNzc9ftiooK7r33XpYsWTLM38o31GFM\nZWVlfO5zn2PTpk08/vjjI3rNofI8j9oVK6g6dIhwQUHX9qadO6ndsIGa5cuH1O8jPUY6+72oqIiH\nHnqIOXPmsHbtWq688kpmz57NJZdcMqLXH4x/vq7g0KFDFMT1286dO9mwYQPLB+m3kT4/Jp19/w//\n8A/ccccdFBUVsXv3bmpqapg7dy7vfe97R/T6A/E8j9o/rqDqtEOET487X5t3UvvHDdRcP8RzfoTH\nSGe/x3zpS19i/vz5GmopEkfxiOKRmLEQjyQilkjEMcZaLAKKR+IpHhndxtT0kLq6ehobe364A+Tn\nh2lsrKKurj4lx+jP2rVrueSSS5gwYQJTp07l1ltvxfO8rv2f/exnmTRpEuFwmHPPPZeNGzeecIyj\nR4/yrne9i+XLl/f5Gu9+97t5//vfz5QpU4bdzpNVX1dHVWNjjw93gHB+PlWNjdTX1aXkGP1JRb/f\nddddzJnjF56/+OKLWbRoEc8+++yw2zxUdXV1NDY29viAB8jPz6exsZG6QfptpM8fTCr6/swzz6So\nqAjwP+ACgQBbt24dUbsHU/9MHVUzGwkX9zpfi/KpmtlI/TNDOOcTcIz+pKLfAZ555hk2bNjAxz72\nMQUJInEUjyge6Uu2xiOJiCWSGY9kaywCikdiFI+MfmNqpEVDQwv5+f6H+5YtEJf0A8Ls2NHC/v0D\nH2Plyhaam3sGCEVFMHu2Hyg0NAx/yb7c3Fy+//3vc+GFF7Jjxw6uuuoqfvzjH3Pbbbfx2GOPsWbN\nGjZv3kxJSQmvv/464XB3O8yMQ4cOcdVVV3HllVdy9913D7sdidbS0EA4P9+/06vjw0DLjh0M1vEt\nK1cS7vkH6+r4cH4+LQ0Nw25fqvu9tbWVF154gU996lPDbvNQNTQ0kB/t+y1btvTIdAPs2LGD/QP0\n/cqVK094TlFREbNnzyY/P5+GEfQ7pK7vv/Od7/DNb36TY8eOMWvWLP7u7/5uRO0eTEtjA+HTo+f8\n0S3Q0eucf2MHzBjknH9jJeE5zXA4bmNuERTPJlyUT8vWzD7nOzs7ufXWW/nZz37G+vXrh91WkWyk\neCQ9FI/0lKp4ZKSxCCQ3HsnWWAQUj4DikWwxppIWntc9sKS5GY4c6bm/rS3Atm0DH+PgwQBtbUN7\njZN1wQUXdN0+7bTTuPnmm3niiSe47bbbCAaDHD16lE2bNnHRRRdx5pln9njurl27WLx4MTfeeCP/\n/M//POw2JEMgLmPaV8cH2toYrOMDBw8yUMf3eI2TlOp+/+QnP8n5559PTU3NsNs8VPHZ6ubmZo70\n6vu2tja2DdD3Bw8epG2AfvdG0O+Qur6//fbbuf3221m3bh33338/JSUlI2r3YAIurl86muF4r3O+\nvQ2atw18jPaDcHyAc95l9jn/gx/8gIULF7JgwQIFCSK9KB5JD8UjPaUqHhlpLALJjUeyNRYBxSOg\neCRbjKmkRTAY6bodHaHVQ1FRhJkzBz7Gli0R+kqw9/UaJ+uNN97gc5/7HPX19bS0tNDR0cGFF14I\nwJIlS/j0pz/Npz71KbZv38773vc+vve971FcXIxzjocffpji4mJuueWWYb9+skTi57n10fGRoiIG\n6/jIiZeiehwrMsRCQn1JZb9/4QtfYOPGjSkrOBY/x7Coj74vKipi5gB939cVkfjjDLWAU39Sfc6f\nf/75PPbYY9x55538+7//+4jaPpCIxfVLbh/nfH4RFM0c+Bj5WyCv1zkfd6wer3GSkt3vu3fv5oc/\n/CH19cMfni6SzRSPpIfikW6pjEdGGotAcuORbI1FQPGI4pHsMaaSFhUVIXbubCI/P8zs2T33tbUd\n4fLLQyxdOvAxJk4MsWZNU9ewzt7HqKgIDbt9//iP/0hlZSW///3vGTduHCtWrOC+++7r2n/rrbdy\n6623cuDAAT7wgQ/wb//2b9x9992YGZ/4xCc4fPgwS5cu5c9//jOh0MDtONmKuyMRqqigaedOf0hm\nr44/0tZG6PLLGazjQxMn0rRmTfewzt7HqKgYdvtS1e933nknjz32GE888USfH9rJUFFRwc6dO8nP\nz2d2r75va2vj8ssvZ+kAfT9x4kTWrFnTNayz9/MrRtDvkNpzPsbzvKTPIw2VVtDUvJNwUT4U9zrn\nj7YRmnM5zBrknN8xkabgGv8YvRw52kaoNHPP+bVr17Jnzx7mz58P+EOQW1tbmTp1Krt27Urp+49I\nJlI80k3xSLdsjUdGGotAcuORbI1FQPGI4pHsMaYKcVZXV1Ja+hTt7U09tre3N1FW9jTV1ZUpOUZ/\nmpubKS4uJhQK8dprr/GTn/yk6z/Tiy++yPPPP4/neYRCoR5LBcUKyvzoRz/izDPP5Nprr+13CF0k\nEqGtrQ3P84hEIrS3t494iP9gKqureaq0lKb29h7bm9rbebqsjMrq6pQcoz+p6Pdvf/vb/Pa3v6Wu\nro4JEyYMu60nq7q6mtLSUtp79Vt7eztlZWVUD9JvI33+YJLd9845fvrTn3LkyBGcc6xdu5Yf//jH\nXHHFFSNq92AqL63mqW2lNDX3Ol+b23l6exmVlw7hnE/AMfqT7H5funQp27dvZ/369axfv567776b\nBQsWsG7dOgUIIigeAcUjfcnWeCQRsUQy45FsjUVA8YjikewxppIWwWCQ5ctrWLRoE+XljzN+/GrK\nyx9n0aJNQ1oaLFHH6M/3vvc9fvOb31BSUsLNN9/Mhz70oa59b7/9NjfffDOlpaXMnDmTU045hS98\n4QtAz7WK77nnHqZPn8511113whs7wC9/+UtCoRD/9E//xJo1aygsLEz6EM5gMEjN8uVsWrSIx8vL\nWT1+PI+Xl7Np0aIhLQ2WqGP0JxX9/tWvfpUdO3Ywe/ZsiouLKS4u5jvf+c6w2zxU/vm6nEWLFlFe\nXs748eMpLy9n0aJFQ1oebKTPH0wq+v7+++/n9NNPp6SkhI985CN85jOf4dOf/vSI2j2YYDBIzfXL\n2eQt4vGt5azeMp7Ht5azyVs0pKXBEnWM/iS73/Py8pg4cWLXTzgc7tomIopHQPFIX7I1HklELJHM\neCRbYxFQPKJ4JHtYNiz7Ymaur9/DzLSsjYhIhjMzvv/97/PhD3+Y8vLydDcnq0Q/B3U5KUUUj4iI\njF5mxooVK7jllltOWF5XRmak8ciYGmkhIiIiIiIiIqOHkhYiIiIiIiIikpGUtBARERERERGRjJTV\nSQszo6OjI93NEBGRfhw/fpxAIKs/ikQUj4iIZDjFI5ktq/8y5eXlvPXWW+luhoiI9OPFF19k8uTJ\n6W6GSFIpHhERyWyKRzJbVictbrrpJm677TZaW1vT3RQREYlz/PhxnnnmGa677jre+973ApCbm5vm\nVokkx0033cRnPvMZxSMiIhkmPh657rrrcM6Rk5OT7mZJL1kdId5xxx1UV1dTXFxMZ2dnupsjIiJR\ngUCAyZMnc/311zN37lxKSkooLi5Od7NEkuKOO+7gXe96l+IREZEMEx+PzJkzh1NPPZVgMJjuZkkv\nlg3rhve3LnrMxo0bWbVqleaTiohkGOccEydOZNmyZRQVFaW7OVlnpOuiy8kZKB5xzvH888/zwgsv\n0NnZiZn+LCIimWTmzJlceeWV5Ofnp7spWWek8ciYSFqAHyx4npeiFmWvJ598kssvvzzdzRgT1Nep\npf5Onfi+DgQCmhaSREpapNZQ4pFIJKKLKAmg9+zUUV+nlvo7deL7OicnR9NCkmik8ciYiRTNjLy8\nvHQ3Y9R76qmnuOKKK9LdjDFBfZ1a6u/UUV/LWBYIBBSPJIDeR1JHfZ1a6u/UUV+PHlldiFNERERE\nRERERi8lLUREREREREQkI2VNTYt0t0FERCQTqaZF6igeERER6duYL8QpIiIiIiIiItlH00NERERE\nREREJCMpaSEiIiIiIiIiGUlJCxERERERERHJSKM6aWFmV5rZa2a22cy+lO72ZBszm2Fmq8xsg5m9\namafiW4vNbM6M3vDzGrNbHy625otzCzHzF42swej99XXSWBm483sT2a2ycw2mtk71NfJYWafjb5/\nvGJmvzGzfPV14pjZf5vZPjN7JW5bv/1rZl+Ofma+ZmY16Wl19lE8kjyKRdJD8UhqKB5JHcUjyZXs\neGTUJi3MLAf4EXAlMB/4sJnNS2+rso4HfNY5dxawEPhUtI9vB+qcc3OAv0TvS2LcBmwEYhVy1dfJ\n8X3gEefcPOBc4DXU1wlnZtOAW4FK59w5QA7wIdTXifRz/M/BeH32r5nNBz6I/5l5JfBjMxu1cUCm\nUDySdIpF0kPxSGooHkkBxSMpkdR4ZDQHKxcDW5xz25xzHvA74D1pblNWcc7tdc6ti95uBjYB04Bl\nwC+iD/sFcF16WphdzGw6sBT4GRBbEkh9nWBmFgYWOef+G8A51+Gca0J9nSy5QMjMcoEQsBv1dcI4\n59YAh3tt7q9/3wP81jnnOee2AVvwP0tlZBSPJJFikdRTPJIaikdSTvFIEiU7HhnNSYtpwI64+zuj\n2yQJzGwmsAB4HpjknNsX3bUPmJSmZmWb/wS+AETitqmvE68COGBmPzezl8zsv8xsHOrrhHPO7QL+\nHXgLPzg44pyrQ32dbP3171T8z8oYfW4mhuKRFFEskjKKR1JD8UiKKB5Jm4TFI6M5aeEGf4gkgpkV\nAfcBtznnjsbvc8459LcYMTO7BtjvnHuZ7qsaPaivEyYXuAD4sXPuAuAYvYYDqq8Tw8wm4GfZZ+J/\nQBWZ2d/HP0Z9nVxD6F/1/cipD1NAsUhqKB5JKcUjKaJ4JP1GGo+M5qTFLmBG3P0Z9MzYSAKYWRA/\nSPiVc+7+6OZ9ZjY5un8KsD9d7csilwLLzKwB+C2wxMx+hfo6GXYCO51zL0Tv/wk/aNirvk64K4AG\n59wh51wH8D/AJaivk62/943en5vTo9tkZBSPJJlikZRSPJI6ikdSR/FIeiQsHhnNSYsXgTPMbKaZ\n5eEX83ggzW3KKmZmwL3ARufcirhdDwA3RG/fANzf+7lycpxzX3HOzXDOVeAXBnrcOfcR1NcJ55zb\nC+wwsznRTVcAG4AHUV8n2nZgoZkVRt9PrsAv7Ka+Tq7+3jceAD5kZnlmVgGcAaxNQ/uyjeKRJFIs\nklqKR1JH8UhKKR5Jj4TFI+aP1BidzOwqYAV+Bdh7nXPfTnOTsoqZVQFPAn+le8jOl/FPqj8ApwLb\ngA84546ko43ZyMzeCfyzc26ZmZWivk44MzsPv8BYHrAV+Bj++4j6OsHM7C78L3EdwEvATUAx6uuE\nMLPfAu8ETsGfL3oH8P/op3/N7CvAx/H/Hrc55x5LQ7OzjuKR5FEskj6KR5JP8UjqKB5JrmTHI6M6\naSEiIiIiIiIi2Ws0Tw8RERERERERkSympIWIiIiIiIiIZCQlLUREREREREQkIylpISIiIiIiIiIZ\nSUkLEREREREREclISlqIiIiIiIiISEZS0kJkDDCzMjN7Ofqzx8x2Rm8fNbMfpagN55nZVal4LRER\nEck8ikdEZDhy090AEUk+59whYAGAmd0JHHXO/UeKm7EAqAQeTfHrioiISAZQPCIiw6GRFiJjkwGY\n2WIzezB6+y4z+4WZPWlm28zsvWb2XTP7q5k9ama50cdVmtlqM3vRzP5sZpNPOLjZ9Wb2ipmtiz42\nCNwNfDB6ReV6MxtnZv9tZs+b2Utmtiz63BvN7P+Z2Soze8PM7ohuH2dmD0eP+YqZfSBVnSUiIiJJ\noXhERAalkRYiEq8CeBdwFvAs8D7n3BfN7H+Aq83sEeCHwLXOuUNm9kHgm8A/9DrO14Aa59weMytx\nznlm9jWg0jn3GQAz+xbwF+fcx81sPPC8ma2MPv+iaBtagRfM7GFgJrDLOXd19PklSesFERERSSfF\nIyLSRUkLEYlxwKPOuU4zexXIcc49Ft33Cv6H9Bz8D++VZgaQA+zu41hPA78wsz8A/xPdZtGfmBrg\nWjP7fPR+PnBqtB21zrnDANEApQp4BPiemX0HeMg599TIf2URERHJMIpHRKQHJS1EJN5xAOdcxMy8\nuO0R/PcLAzY45y4d6CDOuX80s4uBq4F6M6vs56Hvc85tjt9gZu/o9RgDIs65zWa2IHrMb5jZX5xz\nXx/ybyYiIiKjheIREemimhYiEmODP4TXgXIzWwhgZkEzm3/CgcxOd86tdc7dCRwApgNvA8VxD3sM\n+EzccxbEtaPazCaYWSHwHuBpM5sCtDnnfg18D7jgpH9DERERyXSKR0SkB420EBmbXNy/fd2m120A\nF50L+n7gB2YWxn8P+U9gY6/HftfMzsD/wF/pnPurme0Abjezl4FvAV8HVpjZX/ETqG8Cy6Kvuxa4\nDz+4+JVz7iUzqwH+zcwi+Fdg/nFkXSAiIiJppnhERAZlzvV+HxARSR8zuxG/QNat6W6LiIiIjE2K\nR0Qyh6aHiEim6X2FRURERCTVFI+IZAiNtBARERERERGRjKSRFiIiIiIiIiKSkZS0EBEREREREZGM\npKSFiIiIiIiIiGQkJS1EREREREREJCMpaSEiIiIiIiIiGUlJCxERERERERHJSEpaiIiIiIiIiEhG\nUtJCRERERERERDKSkhYiIiIiIiIikpGUtBARERERERGRjKSkhYiIiIiIiIhkJCUtRERERERERCQj\nKWkhIiIiIiIiIhlJSQsRERERERERyUhKWoiIiIiIiIhIRlLSQkQykpn9XzP7errbISIiIiIi6aOk\nhYj0ycwuNLOHzKzRzA6b2QYz+4aZjU9RE1z0R0RERMYgM4uY2axe2+4ys1/F3S8xsxVmtt3MjprZ\nFjP7TzMri+7fZmYt0X2N0dhmetzz32Vmq8zsiJk1pO63E5GhUtJCRE5gZpcCq4A1wJnOuQnAlUAH\ncF4/z8lJRlOScEwREREZvbouaJhZHvAXYB7wN865YuAS4BAyFeQAACAASURBVCBwUdzjr4numwLs\nA34Yd7xm4GfAF5LfdBEZDiUtRMYIM5tqZveZ2X4ze9PMbh3g4d8F/ts596/OuQMAzrkdzrm7nHNP\nRI93o5k9bWb/YWYHgTvNLM/Mvhe92rHXzH5iZgVxbbjGzNZFR248bWbnxO1bYGYvmdnbZvY7oIBo\nYGJmr5rZNXGPDZrZQTPrM4EiIiIiWSv+gsZHgRnAe51zrwE45w44577pnPtz7yc659qB+4D5cdte\ncM79GtAoC5EMpaSFyBhgZgHgQeBlYCrwbmC5mdX08dhxwEL8D/XBXAxsBSYC3wL+FZiNPxpjNjAN\nuCN63AXAvcAngFLgp8AD0QREHnA/8AtgAvBH4G/jXucXwN/H3V8K7HLOrR9CG0VERCS7xEZbXAE8\n6pxrGeTxBmBmIeCDwLNJbJuIJJiSFiJjw0XAKc65bzjnOpxzDfhDIT/Ux2Mn4L837I1tMLPvRkdH\nNJvZV+Meu9s593+ccxGgHT8h8Tnn3BHnXDPw7bjXuBn4afSKhnPO/TL6nEvwkyS5zrnvO+c6nXP3\nAS/Evc6vgaVmVhS9/xHgV4iIiMhYVgrsGeQxBtxvZoeBI/gXbr6X7IaJSOIoaSEyNpwGTI0mHg5H\nP7i/DEyKFqY6Gp2WMR04DETw530C4Jz7YrSuxf8C8bUrdsTdLgdCQH3cazwKnBLXhn/u1Ybp0deZ\nCuzq1ebtRK+MOOd2A08D748WAr0SP5EhIiIi2asTCPbaFgS86O1D+DHEQBzwnmgckw/cCjxhZpMS\n2VARSR4lLUTGhreABufchLifEufc1c654uhPiXNup3PuGPA8PadnxBg955LGr+5xEGgF5se9xnjn\nXElcG77Zqw1Fzrnf418lmdbrtU7rdfzYFJHrgWecc4NdWREREZHR7S2gote2CvwLGwArgb+JTvsY\nVHSk5//iJ0MuS1grRSSplLQQGRvWAkfN7ItmVmhmOWZ2tpld2M/jvwh83My+ZGYTAaKjMGbSzzKk\n0Ski/wWsMLPy6HOmxdXN+C/gk2Z2sfnGmdnV0SkfzwAdZvaZaI2L99Fd9TvmfuAC4DPAL4fZDyIi\nIjJ6/B74l2g8ETCzK4BrgD9F9/8Kf9TnfWZ2ZvQxZWb2FTO7Ku44sZoWZmbvwZ8KuyluWwH+CA4z\ns/xorS0RyRBKWoiMAdGEwjXA+cCbwAHgHqCkn8c/DSwBLgdej5vqsYruZcIcJyYwvgRsAZ4zsyag\nDpgTPWY9fs2LHwGNwGb8qt845zzgfcCN+EM9P0CvQqDOubbotpnA/5xsH4iIiMioczf+hY2n8GOH\n7wB/55zbCOCcO45fjPM1/JijCX+0aCnwXNxxHjSzo9H9Xwc+6pzbFN33TqAFeBh/JZJW4ISVR0Qk\nfcy5Pi+aiohkHDO7A5jtnPtoutsiIiIiIiLJl5vuBoiIDIWZlQIfp+fSpyIiIiIiksU0PUREMp6Z\nfQK/GNcjzrmn0t0eERERERFJDU0PEREREREREZGMlBXTQ8xMmRcREZE+OOds8EdJIigeERER6dtI\n4pGsmR7inNNPCn7uvPPOtLdhrPyor9Xf2fqjvk7dj6Reuv/mY+VH7yPq62z9UX+rr7PxZ6SyJmkh\nIiIiIiIiItlFSQsRERERERERyUhKWshJWbx4cbqbMGaor1NL/Z066msRGSm9j6SO+jq11N+po74e\nPbJi9RAzc9nwe4iIiCSSmeFUiDNlFI+IiIicaKTxiEZaiIiIiIiIiEhGUtJCRERERERERDKSkhYi\nIiIiIiIikpGUtBARERERERGRjKSkhYiIiIiIiIhkJCUtRERERERERCQj5aa7ASIiIiLZwvM86urq\naWhowfMCBIMRZszIo5RGju/YQcDziASDhCoqqKyuJhgMprvJIiIiGc2yYT1xrYsuIiJyopGuiy4n\nx8zcd7/7EIcOVVFQEAags9Oj8alvs4y/ckXVPHJycgBoam/nqdJSapYvV+JCRESy2kjjEU0PERER\nEUmQxsbuhAVA49Y6rvRayfFm8+abeyB6kSWcn09VYyP1dXXpaqqIiMiokNakhZldaWavmdlmM/tS\nH/vDZvagma0zs1fN7MY0NFNERESyWCLjkfz8MB0dsHYtPPUUtG5roDg3nwmtjeS9vB7WrOmRuGhp\naEjeLyYiIpIF0pa0MLMc4EfAlcB84MNmNq/Xwz4FvOqcOx9YDPy7makOh4iIiCREMuKRnBxobYWO\nDggc9/yNzpFz3INIBI4f73pswPMS+euIiIhknXSOtLj4/2fv3qPbvs87z7+/AEGQ4AUUqfuV0MW6\n2bIc+aI4lCynpuI4TZ1pm8lkduZ00m7bnTab0eycTtuz046znZ05M52LtqdNmmabbC7b427aNE0a\nOyadWLYs+SJRsiVRN0oEJVKiJJIgwQtI8Efiu3/8ABK8QyIokNTndQ4PceMPX5IS8eD5Pc/zBa5Y\na5ustQ7wMvDiuMckgNLk5VKgw1o7dB/XKCIiIotb1uMRYyA/373cn3DnVQx78zEmOX8rLWmR0DwL\nERGRaeUyabEGaE673pK8Ld2fAjuMMTeBD4F/dZ/WJiIiIg+GrMYj8XgUGE1adPlD9AzFiWMpKHCH\ncKaSFl0DAwRCoSx8CyIiIotXLlstMtnu43nglLX2WWPMJqDWGPOotbZn/ANfeumlkcsHDhzgwIED\n2VqniIjIgnDkyBGOHDmS62UsNFmNR06e/Jf096+no6MAv/8Ahauq+cnASf6RucCSwmL3QYODRONx\njlVUcLC6OqvfjIiISK5lOx7J2Zanxpi9wEvW2ueT138fSFhr/3PaY/4B+E/W2mPJ6z8Fftdae3Lc\nsbTlqYiIyDja8nRm2Y5HBgcHqa2t44c/jNHU5KGgIMGv/mo+5cNtDH7963gSCRLbthE4eJA91dXa\n7lRERBa92cYjuay0OAlsMcZUAjeBzwGfH/eY68BzwDFjzApgK9B4H9coIiIii1tW4xGfz8cLL+wl\nEIDUSaZPfcodzsmlSzAwAE88AS+8MAffioiIyOKTs6SFtXbIGPNF4DXAC/yltfaCMeY3k/d/Dfgj\n4P8xxpwBDPBvrbWRXK1ZREREFpe5ikeKi0cv9/ZCMAiUlLhJi97euflmREREFqGcbh9qrX0VeHXc\nbV9Lu9wKfOJ+r0tEREQeHHMRj5SUjF4eSVoUF0NbG/RMGIUhIiIiU8jl7iEiIiIii1J6pcVIjiKV\nyVClhYiISMaUtBARERHJsvGVFsBoJqOnBzRAXEREJCNKWoiIiIhkWVERmOSc9AmVFsPD7mwLERER\nmZGSFiIiIiJZ5vVCIOBenlBpAZprISIikiElLURERETmQHo3CDBFz4iIiIhMJ6e7h2ST4zjU1tYR\nDsdwHA8+X4JQKEB19R58Pt9dH6vueC2xSBiPdUgYH4HyEHuerr7rY4mIiMiDqaQEbt9WpYWIiMhs\nLJqkxeHDNXR0VFFQEBy5raUlSn19DYcOHcw42eA4DjXfO0zVhg6CmwpGbo/2tlDzvXoOfvaQEhci\nIiIyI1VaiIiIzN6iaQ+JRNyERVMTXL4Mra3g9weJRKqora3L+Dh1x2upqowQ9HVDz2XovQpAsNhP\nVWWEuuO1c/QdiIiIyGKSylH09UEiAeTnQ+rEhyotREREMrJokhZ+v1thcfs23LwJnZ2jt4fDsYyP\nE4uECRb7wemC2E3ovzVyX7DYTywSzuq6RUREZHFKVVokEhCL4W4nkspkqNJCREQkI4smaZHiSX5H\nicTobY6T+bfpsU7qUvJzYor7RURERKY2aTfIhJ4RERERmc6iS1qk9kS3dvQ2ny8x+YMnkTDJsk3j\nTR5oGLAT7xcRERGZxqRzN1VpISIiclcWTdIiHo8CEystBga6CIUCGR8nUB4i2hsHkzyQZSQD0tUz\nQKA8lK0li4iIyCKmSgsREZHZWzRJi/Lyt4nHo2OSFvF4lIqKY1RX78n4OHuerubtpnKisbTqDDtM\ntDfOsWsV7Hm6OssrFxERkcVo2kqLwUH3Q0RERKa1aLY8PXToILW1ddy8GWNgwENRUYJ9+wJUV2e+\n3SmAz+fj4GcPUffaV4k1teEhQSJaTmDZdg5+tlrbnYqIiEhGfD4oKICBgUkqLcDNZFRU5GRtIiIi\nC8WiSVr4fD5eeGEv0ShcugQrV8ILL9z7sfZ+7ACsj7g3bP0C+MuztlYRERF5MBQXu0mLCZUW4GYy\nlLQQERGZ1qJpD0nxJudnDg/P8kCetIqKhHYMERERkbs3Ye7mpD0jIiIiMhUlLaaSnrTQNqciIiJy\nDybM3UxPWmgHERERkRktuqRFXrLhZWholgdSpYWIiIjMUipH0dub3IwsEBjd6kyVFiIiIjNadEmL\nrFVaGCUtREREZHZS7SFDQxCPA8aMzWSIiIjItJS0mIoqLURERGSWpt32VEkLERGRGSlpMRXNtBAR\nEZFZGr9ZCDDJoAsRERGZyqJLWmimhYiIiMwXqrQQERGZnUWXtEhVWlgLicQsDqSZFiIiIjJL01Za\nxGJZKA0VERFZ3BZt0gJmGQd48kYvK2khIiIi98DvH60CnVBpAaq2EBERmcGiS1rkpeUaZtUiYjyj\niQvNtBAREZF7YMwk3SCT9oyIiIjIZBZd0iJrlRYwOtdClRYiIiJyjybM3VSlhYiISMaUtJiOUdJC\nREREZkeVFiIiIvdu0SUtstYeAqq0EBERkVmbUGlRXOz2jYAqLURERGaQ06SFMeZ5Y8xFY0yDMeZ3\np3jMAWPMaWPMOWPMkZmOqfYQERERuRtzEY+kS1VaxOPgOIDHA4GAe6MqLURERKaVN/ND5oYxxgv8\nKfAccAM4YYz5obX2QtpjyoA/Az5hrW0xxiyd6bhzkrTQIE4REZFFaa7ikXTju0HKy3EzGX19qrQQ\nERGZQS4rLZ4Erlhrm6y1DvAy8OK4x/xT4G+ttS0A1tr2mQ6a1fYQzbQQERFZ7OYkHkk36dzNCT0j\nIiIiMplcJi3WAM1p11uSt6XbApQbY94wxpw0xvzzmQ6q9hARERG5C3MSj6SbdO7mhOmcIiIiMpmc\ntYcANoPH+ICPAD8HBIB3jDHvWmsbxj/wpZdeAiAahc7OA1RWHshC0iL541HSQkREFoAjR45w5MiR\nXC9joZmTeATgwIEDHDhwYPpKi95eSCTcORciIiKLQLbjkVwmLW4A69Kur8M9u5GuGWi31vYD/caY\nt4BHgSmDhBs34Otfd2/L2panmmkhIiILQOpNcsqXv/zl3C1m4ZiTeCRdIODmJBKJSSotrIVYbGw5\nhoiIyAKW7Xgkl2n9k8AWY0ylMSYf+Bzww3GP+XugyhjjNcYEgKeA89MdVFueioiIyF2Yk3gknTFj\nCyuAKXpGREREZLycVVpYa4eMMV8EXgO8wF9aay8YY34zef/XrLUXjTE/Ac4ACeDr1tppgwTNtBAR\nEZFMzVU8Ml5xMXR3T1JpAZprISIiMo1ctodgrX0VeHXcbV8bd/2/Av8102POWdLCWvdUiYiIiCwq\ncxGPjKdKCxERkXuz6KY+zcmWpwB2tgcTERGRB1WqsGIkP5GetFClhYiIyJRyWmkxF+ak0gLcaov0\n6yIiIiIZSuUoYjE3PvH6fFBQAAMD86LSwnEc6o7XEouE8ViHhPERKA+x5+lqfD7FPyIikjtKWkxn\nfNJCRERE5B6kj7Do64PSUtxMxsBAzistHMeh5nuHqdrQQXBTwcjt0d4War5Xz8HPHlLiQkREckbt\nIdNR0kJERESyYNIRFqlMRo6TFnXHa6mqjBAMWOi/CbEWsEMEi/1UVUaoO16b0/WJiMiDbdElLTxp\n39GsKy3GzLRQ0kJERETuzaSbhaQyGTluD4lFwgSL/eB0Q/dl6LkCw4MABIv9xCLhnK5PREQebIsu\naWHMaIuI2kNERERkPpix0sLa+76mFE/qxIxJDwuHJ94vIiKSA4suaQFKWoiIiMj8MulmIakbh4bc\n2RY5kkhVlpq0wWA2MfF+ERGRHFiUSYvUXAvNtBAREZH5wOuFQMC9PKHSAnI61yJQHiLaG2dMWJhM\nWnT1DBAoD+VmYSIiIizSpEXWKi0000JERESyZMLczUl7Ru6/PU9X83ZTOdG+tLM9NkG0N86xaxXs\nebo6Z2sTERFZdFuegtpDREREZP4pLobbt+dfpYXP5+PgZw9Rd+T/I3apEQ8JEuVBAmue5uBnq7Xd\nqYiI5NSiTFqoPURERETmm/laaQFu4mLv/k/Cigb3hnX/CJY8mtM1iYiIwCJNWqjSQkREROabVI4i\nGnV45ZVawo2NOO+/jw8I5eVR/eSTua1qUNwjIiLzkGZaTMek5XQ000JERERmoaQEhocdjh8/zOuv\nv0VbeztdiQRtsRhH6+o4fPgwjpPDeGNM3DPbclUREZHsWJRJi6y1hxgzetZBZxxERERkFoqLobGx\nlv7+CB5PAdFolDN37tA7OIjfWiKRCLW1tblboCotRERkHlJ7yEw8PveFWy/eIiIiMgslJdDZGSYv\nz09vb5xLlz7ADgzgHRhgZzCI3+8nHA7nboHG656wsVaVFiIiMm8sykqLrCctQEkLERERmZXiYkgk\n4wlr/axYsQK8XtpiMfqS0zlz2x5iRltEFPeIiMg8sSiTFqn2kKwkLUwyaaGZFiIiIjILJSXgSZ4M\nGRyE9evXjwQtTR0dMDyc++1FPUpaiIjI/LIokxapSotZz7SAtEoLlUmKiIjIvfP5YPnyEENDcQYH\nIRAIsGLZMgDaYjEid+4QCoVyu8iRkzWKe0REZH5Y1EkLtYeIiIjIfLJ7dzWFheX09cUB2LBhAwBD\niQRdbW1UV1fncnmqtBARkXlHgzhn4lF7iIiIiGRHWZmPvXsP0d1dy7JlYRyPh81LltA/NETl8uVE\no1GWLl2auwWqwlREROaZRZm0yNqWpzBaJqkzDiIiIjJLxcXg9frYuPEFfvu3gb4+2np6+MqJE9ih\nId566y1+8Rd/MXcLTA3i1MkaERGZJ9QeMhO1h4iIiEiWlJS4n3t63J1FCQRYVlLCzuXLYXCQs2fP\n0tHRkbsFqtJCRETmGSUtZqKkhYiIiGRJcbH7eWgI4nHcbUaLi9m/YQPE41hreeutt3K3QFVaiIjI\nPLMokxbp7SHWzvJgmmkhIiIiWZKqtADo7U1eKC5meVERO5I7iZw9e5ZIJHL/FweqtBARkXlnUSYt\nUpUWAInELA+mSgsRERHJklSlBbgtIsBIJuOZdesASCQSHD169D6vLElxj4iIzDOLPmkx6xYRk3bG\nwc42AyIiIiIPsqkqLQBWGMP27dsB+PDDD+ns7LzPqyOtPUSVFiIiMj8syqRFXtqeKLPeQSR1xgFU\nKikiIiKzMl2lBbEYz1RVATmstlClhYiIzDOLMmmR1UqL9KSF5lqIiIjILBQUjJ5cGV9pAbCyuJht\n27YB8MEHH9DV1XV/F6hKCxERmWdymrQwxjxvjLlojGkwxvzuNI97whgzZIzJaOPyOUta6KyDiIjI\nojNX8cjkxxjNUUyotADo7WX//v1AjqotPGqLFRGR+SVnSQtjjBf4U+B5YAfweWPM9ike95+BnwAm\nk2PPyUwLUNJCRERkkZnLeGQqqRzFZJUW9PSwevVqHnroIcCttohGo7N5urtj0nps1RYrIiLzQC4r\nLZ4Erlhrm6y1DvAy8OIkj/tfgb8B2jI98NzNtFDSQkREZJGZs3hkKjNVWgA888wzAAwPD/P222/P\n9ikzN6YtVkkLERHJvVwmLdYAzWnXW5K3jTDGrMENHL6avMlmcuDstoekZUA000JERGSxmbN4ZCoT\nKi2Kity+ERjJZKxZs4YtW7YAcOrUKbq7u2fzlJlLj3t0skZEROaBvJkfMmcyecE/DPyetdYaYwzT\nlGO+9NJLI5e3bDkAHADUHiIiIg+OI0eOcOTIkVwvY6GZs3jkwIEDHDhwYMJjUpUWAwPgOODzeSEQ\ngL6+tEyGW23R0NAwUm3xwgsvZPYdzYZRpYWIiMxOtuORXCYtbgDr0q6vwz27kW4P8LIbH7AU+KQx\nxrHW/nD8wdKDhKYmaGhwL6s9REREHhTj3yR/+ctfzt1iFo45i0emMr4bZMkS3ExGX19azwisXbuW\nTZs2cfXqVU6dOsW+ffsoSf/iuaBKCxERmaVsxyO5bA85CWwxxlQaY/KBzwFjXvyttRuttSFrbQi3\nj/RfThYgjKfdQ0RERCRDcxaPTGXc3E3XhJ4RVyroGxoauj+zLVRpISIi80zOkhbW2iHgi8BrwHng\nr621F4wxv2mM+c3ZHHvOkhaaaSEiIrKozGU8MpVJ5m5OMp3TtW7dOjZu3AhAXV0dPePuzzqdrBER\nkXkml+0hWGtfBV4dd9vXpnjsFzI9blZ3D9FMCxERkUVtruKRqUxbadHXB4kEeEbPKz3zzDM0NjYy\nNDTE8ePH+cQnPjHbJUzNoy1PRURkfslle8icUXuIiIiIzFeBwGhOYkKlRSIBsdiYx2/YsIFQKATA\nyZMn6R3XQpJVRhWmIiIyvyhpMRPjHd2GTEkLERERmSWPx93lFNIqLdLLLyZJSjzzzDMAOI7D8ePH\n53BxqrQQEZH5ZVEmLdLbQ2aftDCj1RZKWoiIiEgWpHIUI/mJ9EEXk8ytqKysZMOGDQCcOHGCvr6+\nuVmYZnmJiMg8syiTFumVFrOeaQGjpZJ68RYREZEsSOUoMq20gNGdROa02sJoy1MREZlfcjqIc65M\n1x7iOA51tbXEwmE8jsOgx8O5zk7yliwhkUjg8/kIhUJUV1fj8yWTFaq0EBERkSy620oLcKst1q9f\nz/Xr1zlx4gQf+9jHCAQC2V3YmFleag8REZHcW/SVFulJC8dxqDl8mO1vvcXH79zhYx0dfPijH9H8\n6qu8+6Mf0d7eTltbG0ePHuXw4cM4TjJJoaSFiIiIZFH6ZiHDw4DPB36/e+MUlRbGmJHZFoODg7zz\nzjvZX1h6pYUqTEVEZB5YlEkLj2d0dmZ6e0hdbS1VkQjBoSF47z1qz50j0t9Pqd/P8M2bHD9yhEQi\ngd/vJxKJUFtbmzygkhYiIiKSPendICPjKSb0jEy0ceNG1q1bB8B7771HbNxOI7NmzOgwTlVaiIjI\nPLAokxbGjFZbpFdaxMJhgl4vnDkDAwOE6+vxOw63enu53ddHx+3bnDlzhqGhIfx+P+FwOHlAzbQQ\nERGR7EnvBpmw7ek0W5qOr7Z49913s7+4VLWFTtaIiMg8sCiTFjB50sLjOO7WIpWVADiJBNy+zTIg\n6PdjrKWrq4sPPviAwcFBtYeIiIjInEivtBgprEhlMqZJWgBs2rSJNWvWAG61RX9/f3YXl4p7rCot\nREQk92ZMWhhj/tgYU2qM8RljfmqMaTfG/PP7sbjZSG17mt4ekkgN1ly9GnbuxOdxv/28jg52BQKU\nJiOI3t5eTp8+raSFiIjIPLFQ45GpTFtp0dMD1k75tcaYkZ1E4vF49qstVGkhIiLzSCaVFgettd3A\nzwNNwCbgd+ZyUdkwWaVFIBQiGo+7V5YtI7R9O8lrOB0dbKmoYOWKFQD09PRw6dIl2tralLQQERHJ\nvQUZj0xl2kqLoSFIxStT2Lx5M6tXrwbcaouBgYHsLU6VFiIiMo9kkrRIjZH+eeBvrLVRYOr0/zwx\nWdJiT3U1b5eXjyQuqnftonz9eqLWcj0vj9WOw9a8PFasWEFhYSGrVq3iG9/4Bu2RqHsAzbQQERHJ\nlQUZj0zF64XUbqUTKi1g2mGcMHa2xcDAAO+99172FqeTNSIiMo9kkrT4kTHmIrAH+KkxZjmQxXT+\n3JisPcTn83Hw0CEu7NvHz5Yt41hFBY/+8i+z7p/8E/Y++igVgQDLOzv5lU2b+N9/7/fwer309/fz\n+s/eJNIZ0Yu3iIhI7izIeGQ66d0gwBQ9I1N76KGHWLVqFQDvvPNO9qotUu0hqrQQEZF5IG+mB1hr\nf88Y88dA1Fo7ZIzpA16c+6XNzmSVFuAmLva+8MKY2w4CRKPw3e9CW5t7461blHzyk/zDT35CfDDB\n2bNn2bZ9mBU753zpIiIiMs5CjUemU1ICd+7cW6UFjFZbvPzyywwMDPD++++zf//+2S9MlRYiIjKP\nZDKI87OAkwwQ/gD4LrB6zlc2S1MlLaYUDMIXvgDJadw0NLDn3Dk+++lPg8eHtZYL5+s58f4cbC0m\nIiIi01qo8ch0ZltpAbB161ZWrlwJuNUW8RlmYWRElRYiIjKPZNIe8ofW2m5jTBXwc8BfAn8+t8ua\nvbtOWoDbXPorvwKbNrnXm5vZ8e67VO97Fm/ygK+9+g+8+eab2GmmeouIiEjWLch4ZDrpO5xaC/j9\no/2tGVRawNjZFv39/bz//vuzX5gqLUREZB7JJGmRetv/88DXrbX/APjmbknZMdlMi4zk58PnPw87\nk30gd+6w+uQpdm/dis/nw8Mwb7zxBj/5yU+UuBAREbl/FmQ8Mp1UpUUiAbEYYMzYTEaGtm3bxvLl\nywG32mJwcHB2C/Noy1MREZk/Mkla3DDG/AXwOeDHxpiCDL8up+6p0iIlLw9+6Zfg8cfd690xShoa\neGzzFsqCRYC7vdjf/d3fMXxPTyAiIiJ3aUHGI9OZtBtkQs/IzNKrLWKxGCdOnJjdwoy2PBURkfkj\nkxf7fwz8BHd/9C5gCQtgX/RZJS0APB741KfgmWcALzgOgcuX+Gf7q1i6dCkAZ86c4eWXX8ZxdCZC\nRERkji3IeGQ6k87dvIdKC4AdO3awbNkyAI4fPz67agtVWoiIyDwyY9LCWtsHtAFVyZuGgCtzuahs\nuOf2kHTGwLPPwr4D7vXhYUpf+zG/WlXFmuTAzoaGBr797W/T398/q/WKiIjI1BZqPDKdbFVawNhq\ni76+Pk6ePHnvCxuZaTGUHLYhIiKSO5nsHvIS8G+B30/elA98Zw7XlBWzrrRI99gTsG27m8QYdgj8\n8If8yu7dbEoO7Gxubuab3/wmPXcZYIiIiEhmFmo8y5CvtAAAIABJREFUMp1JKy1SN8bjcJeVnDt2\n7BipBj127Ni9V4Kmdg8BtYiIiEjOZdIe8o9w90HvA7DW3gBKpv2KeSCrSQuPD1asgJ0Pg89AIkH+\nj3/M5zdsYGdyYOedO3f4xje+QSQSycITioiIyDgLMh6ZTn6++wFplRbp5Rd3eTLE4/Fkp9rCkzbf\nNKGkhYiI5FYmSYu4tTaRumKMKZrD9WRNVtpDUlIv3hUV8OKnoKDAfY6f/YxfCgZ5fM8eADo7O/nG\nN77BrVu3svCkIiIikmZBxiMzSeUoJlRawF3PtQDYuXMnFRUVwCyqLdKTFlZzLUREJLcySVp8zxjz\nNaDMGPMbwE+B/3tulzV7Wa+0SFmxDL7whZGgwnP8OJ9KJNhf5bbY9vb28s1vfpNr165l4YlFREQk\naUHGIzNJ5SiyUWkBbrXF/v37k8fs5dSpU3e/qPT2EFVaiIhIjk2btDDGGOCvgb9NfjwE/IG19k/u\nw9pmJatJC5NeJum4rSK/9muwZIl79+nTfLyjg+efew6AeDzOd77zHS5dupSFJxcREXmwLeR4ZCbZ\nrrQAeOSRRygvLwfg7bffZuhuy0494+IeERGRHMqb+SG8Yq19GKiZ68VkU6o9ZHjYHXxtzCwONtmL\n95IlbuLiu9+FW7dwzp2D995jjdfL8XPnSBjDfzl/nkeefBKfz4fjOPh8PkKhENXV1fh8vsmfS0RE\nRCazIOORmaRXWlgLpqjI3XY9kbinSgsYrbb4wQ9+QE9PD6dOneLJJ5/M/AAaxCkiIvPItJUW1loL\n1Blj7uKVbn5IVVpAFqotxrx4p51xKC6Gf/EvcFavpubdd9n+/vv8+sWL/P769WyMxWg8fpy/OHyY\nkydP0tXVRVtbG0ePHuXw4cP3PtFbRETkAbOQ45GZpCotHMfdMARjoCg5ruMeKy0Adu3axZJkRehd\nV1uo0kJEROaRTGZa7AXeMcY0GmPOJj/OzPXCZiurSQuPF0zygONfvAsKqFu5kqrCQoJ5edDTw0PN\nzVQWFWGB8kSCD0+epLGxEQC/308kEqG2tnaWixIREXmgLMh4ZCaTdoNM6Bm5e+mzLbq7uzl9+nTm\nX6xKCxERmUcySVp8AtgEfBz4dPLjF7Lx5MaY540xF40xDcaY353k/v/JGPOhMeaMMeaYMWZXpsfO\natICRs86THLGIdbcTHD3bnfWBUAsRm9LC4+vXk3A52N4YIDr16+PbIfq9/sJh8NZWJSIiMgDY0HG\nIzNJn7s5krSYMJ3z3uzatYuysjLArbYYzjQgUqWFiIjMI5kkLbon+bgx2yc2xniBPwWeB3YAnzfG\nbB/3sEZgv7V2F/BHwF9kevy8tJMEWd32dJIXb4/juP2n27aNJC6cvj6Kh4d5bOVKvMmBGlevXsWt\ncEXtISIiIndnQcYjM0mvtBgprEhlMmaZtPB6vezbtw+AaDTKBx98kNkXjtk9RPGKiIjkViZJi1NA\nO9CQ/GgHrhljThlj9sziuZ8Erlhrm6y1DvAy8GL6A6y171hro8mr7wFrMz34nFVaTLJfeSI1VNMY\n2LQJvF58xkAkQmFeHsuTE7z7+vq4desWgAZxioiI3J0FGY/MZNpKi76+WQcxu3fvJhgMAnD06NHM\nqi3SKy3UHiIiIjmWSdKiFviktbbCWluBeybiH4DfBr46i+deAzSnXW9J3jaVXwNeyfTgWU9apM46\nTHLGIRAKEY3H3Sv5+bB+PaGCAuLxOAPd3awOhfD7/QCEw2FisRihUCgLixIREXlgLMh4ZCYFBaPV\noRMqLcBNXMxCerVFV1cXH3744cxf5FGlhYiIzB+ZbHn6UWvtr6euWGtrjDH/zVr7G8aY/Fk8t830\ngcaYZ4FfBT421WNeeumlkcsHDhxgxYoDI9fnuj1kT3U1NfX1VEUiBP1+WLuW6pYWTvb0cLa7m20b\nN5IfDHLx4kVisRjd3d1UV1dnYVEiIiKjjhw5wpEjR3K9jLmyIOORAwcOzHBMt7Ciq2uSSgtwMxml\npZkucVK7d+/mrbfeoru7m6NHj/Loo4/iTT+7M2FRqrQQEZF7l+14JJOkRWtyKNXLgAH+MXA72QOa\nmMVz3wDWpV1fh3t2Y4zksKuvA89bazunOlh6kABw+fLo5bkexOnz+Th46BB1tbXEwmE8jkPi8cf5\n+NKl7Aaa+/oo37qVaDSKx+Nh9erVDAwMqEVERESyavyb5C9/+cu5W0z2Lch4JBOppMWklRaznGsB\nkJeXR1VVFa+88gqdnZ2cOXOGxx57bOovUKWFiIjMQrbjkUySFv8U+PfAD5LXjwGfB7y4AcO9Ogls\nMcZUAjeBzyWPO8IYsx74PvDPrLVX7ubg2W8PmXqmBbiJi70vvDB6QyIBf/EXcOuW2zLy679O+DOf\n4Vvf+haJRII33niDX/iFrAw9FxEReRAsyHgkExPmbk46nXN2PvKRj3D06FF6enpGqi08nim6hI3H\n3erdDqvSQuYtx3GoO15LLBLGYx0SxkegPMSep6t1YlBkkZkxaWGtbQO+aIwpSV5Pf/W85xdua+2Q\nMeaLwGu4AcdfWmsvGGN+M3n/14A/BJYAXzXuDhyOtfbJTI6fvnvIXFdaTP54Dxw8CN/+NgwOwhtv\nEPr0p3nooYe4fPkyp0+f5qmnnmJFapvUecRxHGpr6wiHYziOB58vQSgUoLp6j14EREQkJxZqPJKJ\nVI5iJD+RnrTIQqUFjFZbvPrqq0QiEc6ePcujjz469Rd4fG4ApUoLmYccx6Hme4ep2tBBcFPByO3R\n3hZqvlfPwc8eUswqsojMOIjTGPOIMeY0UA/UG2PqjDEPZ+PJrbWvWmu3Wms3W2v/U/K2ryUDBKy1\n/3Ny4NZjyY+MA4T0Sou5nmkxpY0bYcsW9/KpU3DnDtXV1Xg8Hqy11NbWZmFh2eU4DocP1/DWW9tp\na/s4XV0HaGv7OEePbufw4Rpt1SoiIjmxUOORTKQqLQYGwHFwg5hAwL0xS5UW4FZbFCcTIm+99RaJ\nxDRdNakWkYQqLWT+qTteS1VlhGBJwZjbg8V+qioj1B2ffzG2iNy7THYP+Qvgf7PWrrfWrgf+DVnc\nn3yuzNmWp3d7xqG62p2yZS3U1rJs2TI+8pGPAHDlyhWuXMl6lems1NbWEYlUYUyQq1dHT/D4/UEi\nkSpqa+tyu0AREXlQLch4JBOTFlZM6BmZPZ/PR1VVFQAdHR2cO3du6gfP0BYrkkuxSJhgsR8Sg9Bz\nBTpPg+PuShws9hOLhHO8QhHJpkySFgFr7RupK9baI0DRnK0oS+73TIspLV8OySQFDQ1w9SoHDhwg\nP98ddF5bWzv9mY77LByOkZ8f5Nw5aG6Gs2fd8RzgJi7C4VhuFygiIg+qBRmPZGLSuZsTekayY8+e\nPRQWFtLQ0MB/+A//gf/+3/87f/Znf8Yrr7wytppSlRYyj3kSMegLQ8d7EGuBwShEL4B1g1aPkm0i\ni0omSYuwMeYPjDGVxpiQMebfAY1zvbDZSp9pkfX2EJvx7miuZ591h3EC1NRQHAiM7Jl++/btzPZM\nv08cx8Pt26MxUjwOra1j7xcREcmBBRmPZGLSuZtzUGmRcv36da5du8adO3e4cuUKbW1tHD16lMOH\nD48mLu61wlRkLiWGoO0dErePQe81SKSdmRwegP6b7sOM5lmILCaZvAP9ArAcd2r23wLLcPcon9fm\nrD3EWnea9t0oLoZkOSa3b8OZM+zdu5fS5L7rP/vZzxgcHMzCIrMhQeO4EPD69dFqC59v/lSFiIjI\nA2VBxiOZmLQ9JHVjb+/dnyyZRm1tLUVFRRQWFgLQ1NSEtRa/308kEhmdt2WSZ3+0e4jMBzYBkVNw\n6U+g9TUCRUVEY0PgK4GyRyDP/fdMXxNd0T4C5aHcrldEsmra3UOMMXnA9621z96n9WTNnCUtwD3r\n4Mlkt9g0H/0ozrvvUvfhh8S+/GU8VVWUtLdzrqmJfuB3fud3CIVC+Hw+QqEQ1dW52a6pry9ALBYl\nLy9IeTlEIm61xa1bUF7eRSgUyOrzabsqERGZyUKORzJRVDQ6/mrCDiKJBMRi7oOyIBwOEwgEWLdu\nHY2NjcRiMZqbm1m3bh1+v59wODkLQJUWMh9YC9HzcPsNiLeP3Lxnz5PUvLeKqqV5BP0FQAK66on2\nDnCstYuDv1aduzWLSNZN+847uQ1YwhhTZq3tul+LyoY5aw+B5FyLwrv6cgeoaW2l6to1gnl5cPYs\nT69dyw9PneJqPM6KUIhAIIDf76elpYX6+noOHbq/2zVFIuA4eygsrCEvr4qHHw7y/vvuNPOrV6Ns\n2nSM6uqDWXs+bVclIiKZWMjxSCY8Hjcn0ds7ySBOcDMZWUpapNo/1qxZQ3NzM47j0NjYSGdnJ1u3\nbh1tDxmptFDSQnLAWuhthNs/hdjN0dt9pbDiAL4luzm4ddg98XU1jMcGSXR0EciHg0+txWdjQDBn\nyxeR7MqkXKAPOGuMqU1eBrDW2i/N3bJmb84GccI9nXWoq62lyu8nWFbmRiTNzbwei7G2pITO/n6i\n7e00NTWxdevWMSWaL7zwQhYW73Ich9raWsLhMI7jTKjqqKkB8LF370EeeqiO7u4YO3Z4uHgxwZIl\nAQ4cOJjVJMLIdlVF+e4wJTxQtH7MdlV7n8ne9y8iIgvagoxHMlVS4oYHEyotIOs7iAB4vV4efvhh\nzp8/Tzwep7OzkxMnTvD4449jrcWMVFqoPUTus1gL3HodeptGb8sLwLJ9UPHESLWzz+cZGyf2Xce5\n+HXqzlwmdvp38JTtUAWvyCKRSdLib3H7RwEsYJKf57U5bw+5S7FwmGBBAWzaBB9+CMPDhM+fZ0Vp\nKXeKirgyMEBrayvr1q0bqbgYKdHMAsdxOHz4MB0dHRQUjFY1pKo6Pv3pQ1y86H6Pe/b4+Mxn9gLu\nz+5//A+H06fr+OM/PspHP+qhoCBBKBSgunrPrF4AYpEwwU350H0R+m+7NxoPBNa521Vdzfz7dxyH\nutpaYuEwHsch4fMRCIXYk6M2GxERyboFGY9kKn2EBTCx0iJLQqEQLS0t+P1+gsEgTzzxBFevXqW1\ntZV4PE5LSwvf/e53+cUnPe7WLGoPkfvAcRzefK2W4ZvHCZoWPB7LkjIfG7dswLtiHyz9KHj90x8j\nfxU174apWnONYHEelJdCXrEqeEUWgSmTFsYYA3wGd+jVGWvta/dtVVng8bgficQctIfcwwu4J1Vu\nuWQJLF0K7e04AwPQ20uosJCrycxKR0cHgYA7N2LM1mOzVFtbSyQSGZOwAPD7/bS3R/iTP6ll+fIX\nyM+Hn/u50fsTCYfm5hquXasiLy/I5cuwahW0tESpr6/h0KF7r77wWMct/UslLAB6w5BfDnlFGW9X\n5TgONYcPU9XR4SaGkqItLdTU13PwPrfZiIhI9iz0eCRTqRzFXFdaVFdXU19fTyQSwe/3k5eXx9at\nWyktLeXmzZts3LiRq1ev8uOOMPu2F7BybSEma88uMpHTd4cff+P/oshJ4Pf5GaCShPVy7tZOfhy2\n/Pa/+hg+78xxXN3xWqoeWUpw8MZoe0nZLlXw5pBb5V1HOBzDcTz4fNk58SkPnul2D/kKcAgoB/7I\nGPOH92dJ2ZOqtsh6pcU99Hcm0v9jbt8O69bh87g//qJYDF+ykbWrs3PkYdn8zxwOh/H73Qx1a2sr\n7777Lrdu3QKgo8NPU5Nb1bB//9iTO7W1dRQUVFFU5PYFXrvmJoL8/iCRSBW1tXX3vKZEXwv0NbtX\nvP7kObMEdLv7bGe6XVVdbS1VkYibsOjogIYGuHaNYFcXVeEwdT/4wej2JyIistAs+HgkE6kcRSyW\nfMnKz4fk63Y2Ky18Ph+HDh1i3759LFu2jLKyMpYtW8aLL77IX/3VX/H4448DEB9McOnyJT48fZLu\n7u6sPb/ICKcXbrxC+LV/R/HQMImEn7Y2w6U7j/Fe15e47vwitzueyzjWjEXCbht24Sr3hnhkZAvU\nYLGfWCR7FcwyM7fKu4a33tpOW9vH6eo6QFvbxzl6dDuHD9dk9eSsLH7TtYfsB3ZZa4eNMQHgbeD/\nuD/Lyg6vFxxnfsy0CIRCRFtaCPr97sI2bSLU10fL2bPYeJxgYSFD7e109fVhQyHiyXkT2ZL6w2Ct\nJRwOMzg4yKVLl/D5CmlqCpKX51BeDnv3jv26cDhGYWGQ9evdXMDAgLsFamWlm7gIh2P3tqCuegKJ\nVqKxIYIlAViyG/pb3X7E/m7ePP0W15ztHPn+4Rn7EWPhsPtz7e+Hc+fGbA0XBGJNTXDhgpuNKSuD\nYHD0c/plZXxFROajBR+PZCJ1wsBat7CitBQ3kxGPZ7XSAtzExVQzs1588UV27NjB6Z/8NwA6I218\n5St/xvPPf5JHH30Ut/BFZBaGB6DtOLS/AwmHjvY4XV0FnGvewembz9I3tIyPfczdUeduYs2RCt2i\nShhoc+P1ngZ3HoavLOMKXsmO2to6IpEqCgqCDA257yGKisae+Hzhhb0zH0iE6ZMWg9baYQBrbcws\nwFep1A4is01aOI5D3bE3iV16Hw8JEuccAms+dldDffZUV1NTX+9WBCTPnFTv3MnJzk7O9vYS8npp\nuHWL4f5+Oo4eZcMjj1C9b9/sFp4mtc5YLMbg4CDgJjDefrue4uLHyc/38YlPjN11xf3e3WoQtyXE\nzQs0NbnBVUXF6P13pTcMzd9nz8MbqXkzQtWezQS9hVBUiRNr5+/fPk9omeHZZx/F63eHxE/XjzjS\nenPt2qR72XsSCfe0VTTqfkwlEJg+qVFY6L6C3mcqrRORB9yCj0cyMb4bpLQU98W2oyOrlRaZ2LJl\nC+sLfpnGd77O7du3GRzo5wc/+AFnzpyhqKiIW7duTTrQW2RaCQc63oe2t2GoHwvcaoV3zq7h2NXf\noCO2euShQ0Oj55IyjTVHKnQ9+VD2MHR+ANbitJ+l7kY5F657Rh6n4Zxzr6EhRjQa5M4dd4fCRMJ9\n77Bjx8zJKMdx3J1hImE81tHvTKZNWmwzxpxNu74p7bq11u6aw3VlRao9ZDYzLUa25Vx7k+DW5H+u\n0naiw0fvaqiPz+fj4KFDE4ZFfvxf/2t2A5fq62n+m7/B09NDZTDIv1q9Gt+f/zk89xzOzp3Uvf76\nrIZMpgZvdXWN7hQ3OAiRyAC3btWyffvjvPLKYV5/fWwA4vO5bRUeDzz8MJw65SaBLlyAPXtg2bK7\nbLvovw3XXgY7jC8/n4NfOEzdmavJ7aocLp5fzr7119lWWY639zLkPw7GO20/YsLnc+tpk+0uLF3q\ntuAMDEA8TqKwEPbtcxMWXV3u5+7uiQmOWMz9uHmTSeXnT0xkpF8uKXF/UFO4l2GhqdK6jg43U52S\njZkiIiILxIKPRzKR3po5UlgxYTrn/eMvKGL7tu0sW7aMD04F6Ojq4zvf+Q6O47Bt2zZWrFiBMSZn\n27TLApIYhs7TcOdNcNwEXG8vfHB5DSdbnuP1q2FisdUUFLihG7gFRql/TqlYdCaB8hDR3haCxX7w\nBaF0K07HeWreu87joUbWfuQp1lbOfDJM7p3jwJUrUF8Pr73mITYuL9HRAadOORQV1ZFIXJ70ZNzI\ne68NHQQ3pc2q0+/sgTZd0mL7fVvFHMnGTIuRbTkDhdCWvNEm7mmoj8/nY+9UW5j+wi/gKyyk68YN\nVnV34/N6oa8P5/vfp+Y//keqVqwguHTpyMPvdshkavDW+fPnk2vJp7c3SDR6AoDy8h6iUfcPeXoA\nEgoFaGmJ4vcHKSqCbdvcP0SDgw61tW/yxBPXOHz4SGZn/wej0PRdGI6719d9Bl/ZVvY+s3XkIQnz\nZ+xc5YGeRhjqh57LULoNMFPuKBIIhYi+/vrobtyhkPvLLyqiy+slsH//2Omi4KZ7u7vHJjLGXx7f\nazc4CG1t7sdkPB731NgkSQ0nEKDmW9+iqqvrroaFppfWpVNpnYg8QBZ8PJKJ9EqLkcKKVCajt9dN\ntN/PIpPkLK+lFUv5zV//Vf7P//In9Pf3k5eXx8WLF2lvb2f79u1ztk27LALWQvQc3H7DnS8BDA3D\n5WvL+Nn5n6N9cCtgWL36NsZEKSkJcuqU+6UDA+7/iYGBLkKhQEZPt+fpamq+V+/G7cV+KFhJXdMl\nHg8NEh3IZ1PpIKmNhzScM3uGhuDqVff9wcWLbrjscpNNXq97PnFwEDo6HC5cqMHaKjZtWkFXl9sK\nn34ybuS9V7Ef4h2ABX+FfmcPuCmTFtbapvu4jjmRjaSFuy2nH2z6QdzLd7st50wqKyv5oKuLa2Vl\nJD79aTy1tdSdOkVVayvBtjZYs8YdJuHzEfT7qYpEqKutnToRksbn8/GlL32J8+fP4zgOfv8KOjry\nKSwsZ+nSYtrbW2hvL2Pp0qVjAhA32VFDJFKF3x9k2TJYu9bh7bf/HgjR3v4snZ1ejJnh7P9Qv5uw\nSGbYWXUQyh6ZsE6PdSCwzv0jNRgd3VkkmbiYrB9xzyOPUNPbS9XQEME1a9yGOSAaj3OsooKD1dUT\nfyAej5tYKCuDDRsm3m+tW3UxVVKjq8vtlUmXSLi3d3W5rSpp6hoaqLp2jWBhIRQUuK0oGzYQDASm\n/T2GwzH8/iDd3e4sEYDNm91DzGqmiIjIArEY4pFMTLpZSOpGx3FPPY/bAWxOeUZDxMICH6tWrWL3\n7t1cvnyZwcFB2tvbuXDhAjt37sz6Nu2ywFkLPVfg9k+h362CtUBrRxm1Hx6gqWsXFg8ejztL7emn\n9/DVr9bQ2loFyVNQ8TjE41EqKo5RXX0wo6f1+Xwc/Owht60gWcF74dZS1m6Ls2llHl4n4g6AL1oP\nZD+Of5AMD0M47I6Su3hxtEImJT8fHnssQFtblJUrgyM7Or75Zh1DQ1WAn/Z2H11dbiiefjKuoO8c\nwTU3oP0WDCczIP4KnMBDXGi4zunwtxjouKyWkQfMdJUWC15qPsNs2kNG3iQbz+iO8InhifdnQSgU\n4oMPPmDQcbhZXs7aL36R2L/5NwRv3nRfAG7ccM/yf+QjUFBA0O8ndhdBQnt7Oxs2bGDt2g30979I\nPH6ckpIgHk8d1g5TX19PUVERpaWllJaWcubMGT75yU9y6NDBMTMViosvsnz5PgoKttHR4aW5Gdav\nn+bsf8KBa3/lDkUCd6/tZU9Puka3H9FAcCd0fghDfTg9rdSduUFsuIQrkcCEP1K+t9/m4N691IXD\nxHbtwpOfP9J6cfBe+2yNcZMfRUWwevXkjxkcnJjISE9u9PSMtKDEOjsJ5uW5gafjuPe1t+Ns2cKF\nnh5Of+tbDFy+PKFlJBLxcO4ctLePPm00Clu3OnR21nHu3OSldfORZnPMP/qdiMwfeXnu6KT+/kkq\nLcDNZNzPpMW4AeSO47B06VKCwSD19fV0dXXR3t7OlStX2Lx587Q7AdxLe6QsUH3X4dbr7uek3ngR\nb17YT13jHhLJtx4bNsCnPgXLlwP4OHToIDU1dVy6FGN42ENBQYJ9+wJUV99dG6zP55twFn7tpnaI\n1Lkn0OK3R5IWkN04frFLJNy5dvX1bpv4+NYPnw8eegh27oQtWwD2cPjw6IlPjwcCgRhlZX56e69T\nWrqJs2fhqaegIH+IdSXXyb/xd3jiH0LZ2CyI09dGzZFzVD0cZOX2cio3q83nQbOokxbZqLQY3XbT\nuC/g1hmz5Wmm23JmorKycuRyU1MTa9euxbN5sxvFNDS4b4QHB93kxaZNQNoQygykzoI0N8PKlRtJ\nJN5k69YA+fnbOXfuHNZaent76e3t5ebNmzQ1NQGwevVq1qxZw3PPrWXNmjV861sJKip2UlfnZlav\nXYNVqxyammrp7Axz+XI94XCdOxvjuZ/Dd/PvRrc2LXvYrbKYwph+xCW7cdpPU/PeRaq2DOEvKOCh\nDXtYW9k2+kfqo7+M79IlfF4vez/3Ofj0p+/uhz4b+fmwbJn7MZnh4ZEWFE9fn5twisfdiLSry+3Z\nq6mhqrCQlWvXUpmcNxJtaeEHJ+phxyHefTdBX597OGPcHEg87vDqqzUUFVWxdu3kpXXz7Q+3ZnPM\nPw/a70RvmmQhKClxXyImVFqAm8lIaxOdc2mVFiSGRv6f+Hw+Hn74YU6fPk1fXx/Nzc20tbWxZMmS\nkfvTZ2M5jkPN4cNUdXTcVXukLDD9t+D2z6D78shNQ/j5oOVjvFa3F2c4H3DPBR08CLt2je128vl8\nfOpTe7l82T0588gjkI1uo4TxgcmD/CVu0iLVopx+v0wpkXArfevr4fx5RmLSlLw8N0Gxc6ebsMjP\nT7/XN+HEZyBwhUceeYiCgk1cveqlNP8267yn2LzkDD5PPwXcJkZyPlx+KRSsgng7dfWXqdrUT9Az\nSGRo9EnUMvLgUNJiBmPeRHt8btVAcsvTrp4BAuXZ25Y0GAyyZMkSOjs7aWpqoqqqyh0yWVQEjz4K\np0+7b4I7OkaSFom7eKFvbGxkYADa28uprAxSVuZj1SowZimPPfYY7e3tdHd309PTQyKRwOPxMDAw\nQGNjI42NjSPH+eCDNny+DgoK1tDVtYa8vKUcPfoVEokO8vIKSCRitLW10dLSTP/V7/GZfWvxerxQ\nHIK1n5m2J3d8P2JdSylV2zz4vXC9fYhNW3qhv5VgQZCqDR3UffNP2espcX/Z+/ff+w9/Lni9sGQJ\nLFlCorJypG0FgI4O6o4edVta+vqItLZCPE4cP7ev+SlsjvCTy7WUlZXT3R1lzZoglZVu3uqdd+pw\nnCq6uvyUlvpwHDe7PZ/nXKTP5kgk3GRXXh7k58/dmlVFML3038nwsNsxldpebr7+O7pXetMkC0Vx\nMdy5M02lxf3kSfs/YZ2Rgd5+v5+8vDx27drFiRMnaGxsZHBwkN27d48M+06fjVVXW+vunFZQ4FYa\ner3g8dx1m6vMU/GIO7MiOrrlvDV5NPc/xd8f+xgdUXcehTHwxBPw8Y9PXzAUDI7OS8+GkTje4+7c\nR2LIbfk23qzH8YuFte6OgefOuYmK8ZsXeb21WGbIAAAgAElEQVTu25CHH4atWyG5KeKk3O2V96Zd\nTxBpX0YZH7DNd4plRTcoLwBfMk+RMPkE1jxN1N9GsMxNhFK4ktjwDYJFPQwMDuNLRKH7Io5/A3Xn\nW4j1dHKl4zKxSFjtIovYjEkLY0wV8O+ByrTHW2vtxjlcV1Zkoz1kzJvo1At4wiHaG+fYtQoOfnaS\neQmzUFlZSWdnJ9evX2d4eNgdMtnS4m6TWlHh/hWPxaC/ny5jCIQy+2M7NDTE9evXaWyE0lL3V/eJ\nT4S4fNkNQILBIMGge7Y1kUjQ2dnJ5s2bqays5MaNG7S1tWGTL0aDg710ddXjOPXcugWx2BXy89so\nKyvH7y/F643hOA5bgrcIDl2h8eogW3ZVwYbPjT1zM4nx/YiXw+dYsWMzvqEomyp9eId7oPsSAMH4\nELFID6z7COzcB8VjXwXn03ZJY36PABUVxJYtIxiLMTAwgHfYEvnxO3TFfCS8hVT6Aqx0+vnkL/0L\nLqz8IV2eT5BXsJyVK6GiIkZ3tx9jruPxbOL4cSgudigcqGW5DVN3uZ6CcB3569YRoZzm5sGcv2lP\nzeYYHnZ3oEnP1Pt8QRobY7S3j3bkjP8oLnY/Z7rsB62K4F6EwzF8viBXr7rBydKl7pkSWHzzUsa8\naUoTyMuj4MQJvv2lL7Fp61ZVX8xzCzkeyYTjOFy+XMepUzHy8z0MDibYsDzBkkuXiEejeDo7STz0\n0P37N2rSKy2ckYHekUgEv9+P3++ntLSUoaEh8vPz6e3t5datWyxZsmTMbKyCcNh97btzx60p93jc\nPzjLlxNcsmSkzVXVUAuM0+PuBhI5BTa5u4fx0Ot7jB/XPcOFhtKRh65ZAz//87Bq1cyHLU1+2XQ7\n1N+NkTh+5e3Rge3DcaID3jmJ4xfqv2Nr3c3zUomK8T9/jwc2bnTjhG3b3CLwu36CWDNPrT1JR/wN\n8vO8XE92UQ8NGSKDG2nq3sz2J4d57uBT7u4heclhnBg8/jIGvENc74qyacMSnJ6b1Py0jqodAYKr\ny9gYSVC5qU3tIotYJpUWfwkcAk6RmkC5QGSj0mLMm+grt/H0x0nk+Qls38fBz2b/D1BlZSWnT59m\ncHCQ1tZW9lRXU1Nf7wbcFRXu1Bsgevs2x3bunHzI5CSam5vp6Bjizh3YuXMjO3fCZz5TzeHDowFI\niuM4rF+/nt/6rd8a+f7i8Tg3b97kxo0bDAwc4f33IS9vCI8HhoYiWBunt7eV7u5mBgdj3L54lS2V\nHbR6vdxsj/Hh4HMUXv4pJSUlFBcXj/lcVFSESau+SO9H9FiH0OYuNyve0wDxNkgMk7AJOlvCNMfi\nvNsZw3PjBmV//zM27diLt6QSx7eCmp/82C0l2zT6lzVXf8zG/B6TP2uP10usYhkX7/SypC9IN+DB\noWDIodDTzeMFXTzf9wOeDwzT2Pgmd2J59BWUUtDfS3+okw4e4uZAG12JEmz9Vzgw1EHAU0BBQYzV\nvltcCB+hpuARlu7/fbxe93vN1Zv21B7rV65MLC10HIhGPaQV80wpP3/mxEZREbz9dt2EhAUsjiqC\nbAVEXV0eTp8ePYPS1ub+blIFQanf2WIQS71p6ulx+9kSCZzly6m5coWqgQE29vYSWrECUPXFPLdg\n45GZpBKtDQ1VxGJB+vuhtdXh/N/8J/6XpjPsXFeBJxKBtrb79280vdIi2R5y6NAhamtrCYfDOI6D\nz+dj9+7dRJPvcC5evAhAQUEBwWCQRCLB0yUl2MFBTEOD+w5leBintZXa8+cJOw4Xyso4d+cOnc3N\n/HZFBUvTqhL1//Huzfmb5qF+aD8G7e+NVB4DDJc8zLvhZ3njeMXIycLCQnjuOXcUW6Yb3yTPn42M\nBZvthjkjcfzPvkOs4RYeEiQqAgRWP5n1OH6hVfVZC7duua0f9fXQ2Tn2fmPcDfl27oTt290Z8ndt\nqM+dURc5BfF2HtswzLutV+nvX088sZzzrY8RufkYayoN5eVv89zBg5MPVG0PsnlDBZv2rMHbd4UT\npxqo2hwn6HOgP4YdXgNAwJ9HQfcJvv0/vsSmzVs1rHMRySRp0WWtfXXOVzIHspG0gLQ30ZuG3UE+\nvmLYPjeljOlzLcLhMGvXruVgsrwy1tiI5/JlEkNDBJYtu6s/fleuNNLQAGCoqKikupopA5D0XtQU\nv99PKBQiFArx1FNPcfhwDa2tu/D5eqmvv83wcCvWdpKXF+WRSj9Pb3AnR8bilp9eKmOH7xpwbdK1\neTweioqKKCkpGZPMKCkpofVOhJ4VfeTn55NfshVTupXhwW7OHf8pdMbpjhkGAgEYHKT7WhNtd26z\nd+9e6j5spKrsGsGBQhguBV8p5JcRLCoZ0/t2v6oxfD7f6O8xHMYOOBwfDBIfqMC3ZiOD8W788W6C\n+f2sDPYTMP3k57lvGr1eL1u2rGVL8lhl0dvEYmex9iwDw3C+tYHHo614PSXEPQGGhroIf9BPV1+I\nRzz9/NipZfMTL1Ba6tBz/Rixugt8vaGWbY9uuW/Zf58vQVsbtLa610tL3eFbjuOOaSkoSLB2rfum\nua8vfbussQYH3Y/xL6zjvf9+jFgsiM/nJjrWrh09w7OQqwiyFRCdPQvvvZeYUPJ582ZqeJb7O1ss\nPJ2dbkSWtl1xXUMDVb29BINBOpM/S2d4mAvXr0NdHf9vQwPrH310QZwhe4As2HhkJql2rZKSILdv\nu28m2htq+ZTTj9+spLOzk4qVbi/+fWurSK+0SM7ycku9xz5nV1cXt2/f5vLlywwnA66BgQEGBgaI\nRqPc8ni4fOMG6x2H9X4/q5cu5a8bG4k4DgUeDz0dHZz5/vcp6e7mm+XlHNq/H19ZGRgz4XtdqGex\n75dMXyPu6ec4PAgd70HbMRhOG5JYsoVw/8f50Y9WEYmM3vzYY27CIr0zNhOpSovhYbcjKr1D6l75\nfD72HvgMrEoOB137IpQ/NvsDjzOmqm9gwC31DgTm3b/jO3fcior6erfjPJ0x7nD9hx92ExXpY3Uy\nZhPQ2+gmKrovjdmB0ev18dTBT/F2fT7HP1jCrU7DkiWnJgxdHT9QtaDiIUp8R/H6/FC2ixgd/P/s\nvWlwHOeZ5/nLzMq6UIVCFW7iIAEQPCVREmWJsiCZkkVakuVj+5x1rGd2emJiNrp3ezS7sR+mYyPG\nsx82omNip9Xd9hzqme7w9HiibUu9tlvWQUqWRIoWJBGUxFMEiPsG6j6ysvLcD29VoQoESVAE1ZZH\n/4giUVlvZr6ZlfW+z/t//s/zRMJFcCx0vYRqJzBXT3PsdJKh3RL9sTx9O8vOiM/VF78W2Axp8YYk\nSf8G+Fugmr3Gdd0zt61XW4RKeMitkhZrByxTjJZ22+qlRyIRYrEYyWSSqakpHn74YfHDrRgJvb1C\nX++5uXQkJ05Mks9DKNTBo48GaWoS2zcyQG4EQXaIxDqOozE/78WyItxzTzP7+2TuiZwFN4ZhOrw3\n2YM/EqW1tZV8Pk9xfZlQRDhKLpcjt34FBcQXxgllpgkHxfV6vV5SqRTZmWWMIrx1qZHR8AE6NYm2\nkEHEyHJlYhotlyKyzSO8AKUElBI4rkMylSFbdDn14SkunXuH5PIs//hoK7GBGs/OusFtq4gNVVW5\n/4mn+PBDePNNWO7chWGcxCcreFqj9PZFicXEY5XWdYIPPgj33w/JpJhVkklIJgnPxsle1vF4/AQC\n0OBJsb3RwbYzGEYCr6RRWFmgw1mhqDTw8VQazWzFE/w5T1hFQmqA4FyK+7d9co/dzU64HR1Bnn8+\nA0TweGDfvrWYVl1P88gjwbqEW4YhoqAKBWG0VMiMa73KkUtVOI5c7qd4jY6Kibdi/HxWVQRXhTk4\nzk3FhhsGvPQSfPghhMNBkskMO3ZEyOdFvpTlZSH/NM00fX2fxKXyK4Z0Gt58E+edd6jLaKsoaLpO\nxHUhncYtFDAjEY4tLjIERLxeJufm6Nu27VfWQ/bfKT6z9siNUAmhq01it3x2Esvrw5Qb0PU4rKzg\nlErMmgorTpAXx15lZNJPT4+XGEmM2dmtXQCtU1pshMrx29vbaW1tJZfLkclkSKfTZLNZZFlG9XhI\npFIUZJlLrsvY4iJzlkVMUQhYFrbfj1Ms0i1JJBMJjv/iFzzV2YkZDnO8UGDSNLkwOsp7V66Q+ugj\n/iAW+1yNcQ1U5wifT6jKdB1CISKhEEMrK4wcPy6UnzdDfjtlh93qCTBr8qo09JANPs7LJ7Zz6dLa\n5vZ2EQrS0/PJriFSI5DMZreGtACE86oCc4tiT9ZBm5wkYtv1JHm5Gl0kFEIzDMydOzn2/PMMZbNb\nTyxdB/H4GlFRw99X0dMjFBX799/CPTcykPpAvIx199jXDLF7oekAHjXE4X7IS3D6tFDk3Ggpsj7n\nnewJQKALvZBiJq0xsD3K++emGerOEJG9pAwbXBvTgktjMxAf4QfPjtG788DnyovPKDaz8j2EKPR5\n37rtj259d7YWFaXFreS0qD9g2Yh3HXBKoNye0mM7duwgmUxW81oolQsB4QY9c0Zc1NTUmlv0Okin\ndd59dx6Arq4+hoZuvY+VxDq7d8Pqqs709En6umy+uO0DVDmA6wY4n7uDxs4QTz/ySJUYMU2zWqGk\nQlRs9LdWrqMUbe/ntUsrPL63SDjowTAM0isrpLM6L38ok1VCxMdmODcm+iXLMm1ndB7a3cDubRBU\nLXxyCVW2WFxaxDItZFkmImkYkz/hy71pkqMhmnbfg+xtAk+IUMBLh/0+3/mjf06sY4DExEl+6wGV\nL9y1RyQU5eZZW9cVE8Ubb6yx2tH+I/yycIF/0pJkoMtX5cAypRKnmps5+tRTIonDuiDQnd/+Ni//\nmxfRF/bS5FjktDj5wiqykSXoX6S3s5H5+Sy67uI18nTrk5Smn+Wrnmkifj+GrxE3lId5LxFVZSid\nZuTHP+bQN74hZg75+gv6m/X2uy5kMgfxeI5RLA6xb1+kSlhoWhwn8T3UK1HefHa4blJualKr5Nr1\n4Dgi234tiZHPC2WHYQjpo+uKOuIHD4rL+6yqCKphDpq2VlFo1y7o7LxhCeSlJfjxj9eevzvuOEhX\n1zEUZYhsNkI6LYaV+fkMe/ee4siRa1f5+ZVHPg8nTwpryLYJNjWRyWaJbNsGO3aA1ysq+iQS6LqO\n6vczMjrKUCZDRFEgGERqbgY+Ra/259gMPrP2yI1QIVIjEeGTsCxQTJNMEab1FrBnCIUc0tMT2HaE\nZsVHrMmH3tHNq8//DV/nLI8P7a3aC1uykK8jLTauVFabnFOW5Wp+rN7eXnRd5869e+n74AP+bmoK\nT6GA1dxMankZSZJYtW0yqkq0pYXs9DQZ0yTiugxnszzU2MhzV66QsCz8sozm9XJ2aYlwJsNfRaNC\njVG+rl81L/bfJ7TJSTGGnTtHnewBiJQ/Hzl5kqHZWSKVCdbn21jVYpS49Op/wr/8Bl43jyvLqE1R\nOvc/CB1HGT4/yFsnpKoy0ueDRx8VvpYbmBHXRS1pkcmIfBhbAtkDngYRrmBuUZbPWuRyyB9+KOJg\naz0privmpHweOZ1m5F//a4amp4nUxrUGg0QCAYaWl2+KWLrRs55MCtvz/HnhlFiPrq41oiISufrz\nTcGxIXdZqCry4/XXLqsQ2S/IimDPVc7eyiNYLIriejdK6FmX8241xPY2GTXQzcDBbpTSIpq+TKRB\nBsfCLSUwl94Wyos9CpFtYSZTc/QNbNuUDf+rlBfvcwjckLRwXffwp9CP24KtCg+pwlPjebS020pa\nnDlzBtM0WVhYoKeWru7vFxdm22LRsgnS4oUXpjEMMYh84xv9m05muBl0dUF//xGyyTPsD76JKguG\naLSwi4V8mObmGEdq8m6oqko0Gq2WRrsWbNuuEhip1O9wZvh15hITWKUC7w9PshxvwHaCyNvqR1nH\ncUhnskwtyoy1VkIAJDKpNG5JpykIYb9DJi+jKCVCPhfJypOaP0dzrBnHdZibXyBYgogWITdxhcf6\nkmSSKsPDaQ4dOoQiK4SCniqx0dq1C1mWwUixd0cUVXaqg9u9Dx5hakrl9dfrJ4xIBA4fVtn3fz3D\nB68fZ2bdhHP0OsaVqqr8b//n09XKGLkVBU3zE42G6em/D9mySJ+6iJNuRC1mKSUdZCtFUPagOBaB\nYhLVSsGYsDQigDY9LeQIkiSIi/IkWk0SUfP3yLvvMjQ3JybcGsXRtRZ3w8MwM6Ny6NBRYISODlHN\nQ5YNQrM/4/diAWLJtUnuZidlWV7rZgWPPRbk5MkMPl8En0/we4WCKNvV0fHZVRHIui4uZmZGsDUg\nSiCXia2NSiC7Lrz3Hhw7tjYW9vXBb/yGit8vFFPj4xqzszKW5RAOBz+7iUqLRfjlL8VDV3MvDj71\nFMdGRxkyjGpOGScaRVcUZiSJgY4ORj/6iIgsixtWKOAahjD6Y7EbEkKf49PBZ9keuREqRKrPBw88\nIOTbmYQKOVjx96JLOdKpLN60QlB18XptgrlFpLf+Nf8kv4Tb0MbUxUkG7twJbBHZVhcesrH3Z31y\nzgpKpRLNzc38TkcHalMTX/jGNzgdibDiulx66SX0bBZbkogGg8iyjOv3k1UUso7DqmXx+zMzrBYK\nNHs8NHk86LKMd3mZiGWRzGY5/uKLPFWuymUGApyKx3lldJRTly5x5e23eczj4Zt79qBeg8T5dSU2\n5GxWVJpbXyu98nk+j/bRR0Q0TWRgBsGSBQLVhbN26m3M/gCXjz/LLiuBvypb9pOb8/PfTmukOvtI\nJtcWoHfeKcqYboUqorFGELFVFUSqUBu3nrQoleDUKXjnHZz5+bX73d4uKshV5KL5PI4so6VSRDwe\nsV+pVBefEQG0K1cYefVVYWc1Ngr7KxAAv7/ud30tYmPpyhz/8cULKHc+w/Kyim2bpCaO40tNojom\ngSaV7oN9fO0fHaGt7RaedX21rKr4SNzTWgS3QfReaLrjumulWsdUOi1u2fVQGzISjPURU0+WE3UC\naj9yaBG8CXQtjer1M3IpztCODBFFBi2NpDeBNkvE42VoW4qRN3/EoUe/KZzSNYSKaZoiEej2BJGB\ntf7H09P8+//nBfbuP1Bn639OZHw62Ez1kCZEtu5KPck3gf/bdd3bo63aQmw5aaHULHRsDYht0YHr\nUZvXYmpqqp608PlEiMjkpCAtbhCmEo/Dm2+KDIfRqMKXv9y7pX31+6Gz3eG+x2MEaCcYTDGj92M1\n7uXhA1fnxtgsFEWpemy6u7u58847xQdnz/K9dxdZ9WrQ34/T3Y2u6xSLxerL4/HgDQfIFy8QCoiH\nIK8ZmKaXtA6FosNLF8LsiObY12HQ4LXQikVM0xTH0sUxIh6bRjLsbBLGmmlLrI7maOvczsWPJzCK\nFm42TjbQSGLmNI/tyhOMh6vExuTMNH/w7ReYzRwAHGRZpbOzj9/7vSM88IBajvBR6wzKCrN78sXn\nrsvs1paQGu7T2X3y5FplEkXB39PKtNNCsbGLk+Eh2hMjzEjzbAtl8WgJwoEa9Q4gVxbAriu8+Npa\nzgfTthlZnkBzU8iSw6XJWcjJHPRGUWVZGDzlV8TjQVteFm5Cv59Ewc+FVwO0SX4amv38g3/cjS/i\nh0CA4TfeYO9ieK3fZWxmUr6RF/HIkYNcuHCMZHKI3t5INcnk+HiG3t7PqIpgehrnvfeudpfk80JS\n4vVeVQJZ0+CnP4XLouAOsgyHD8PQUMUTtvYc7dsnxAkg1uo3Mh5+pWAY8O67wnDUa2KtBwbgscdQ\nu7o4um6RMrpjB+3xOHt270ZRFOR0WlhNhQJ6Po/q9ws3X0yM8xsRQp/j08Vn2R65Efr6gszNCaJV\nVYVDwHdXH5Ercziai0eNMJULYkW2E7aSqPoscaeRmJsg5tqQX0Q9m4IAsHONuLglsm0TSovr5sY6\ncAD1P/9n0a67mwf/6T8FWWYuGGS1rE83DINMJsO4JMHsLCXLQvb5SAGGLLPoOMyaJpptQy5H2nEI\nKgpGIsGgz0eTx8NfLS2RtCyyXi+XJidpyec5o6pMzc7yzAMPoDY0EFQUtPff51/+4R/SuXPnpoiN\nzxwWF8UcUSEsmpvFwG6aVU+/I8six0/NHI9liayXuRyEdLaFTrP8N3/HzkwGv6oCfsj3YDTcw9yM\nj6b5FO9uP07L4FO0tMBXvyqI8K1CMLimNtqqCiJVqBEoLm5NeIhlCTXfiRPV+xmMRskoCpHdu69K\nBpHWdYL33Yf+wQcigVSFzCgW64klXUcbGxPE0vqEEx4PEb8fbXmZkZERhsbGiITDlAomq1k/K0kP\n2ZyP7VaSVzPHifYfITv8LL8pJejb5qetTdzfTHKOt79f/6xvhsgzSwUuvfqXqMvvEiC+pr4Z6EdR\nG6DpLpErJLCJMjHcPGlRi/XhIgAOKrobZKbUzMDgTkaXhok0KIALroNr5iA3DpQJossz0D4GkixU\nOJ4QeEKMjHzMUOsVIkoD6F5Q/JhugHff+4Df3ZvD9Ol0bxdO4xupNjaj2Phc1bE5bCY85C+Bc8Bv\nAxLwbeCvgN+4jf3aEmxFydP6A65TWtwmNDY20tzcTCKRYHJykocffri+weCgIC1SKTGgtbRc81iv\nvgqJhCAtHnmkG5/Pe822nwiOzQMdPyS3uILHM8gXvnKQ+7uevi35PrBteOMN+qJR5opFfF1dyLJM\nMBgkWE5prOs6jzzyCEeOHOHlH/5bdm9bwifb6LqOpmmkszqvfSxhqS2s5g0WckLD6M0q7FCjpONz\nqLgE1BJLqzoB1SWTdZFlGVmWySUMLC2O1yzQ5vOwI5DBLazy9XtyBAJeDDvN9AUL2+nhyvglHu3W\n+UliiUBjP63tCpHox5x5533uv/cPQQ4JT1b5Xl2L2b3RgLhRZZL+/k4mFj7mFe4kOvgVMrlpFi0X\nKRIj2Ckz+MADwlNfzoTphELw1FOYmQzHT51icmYGs1hENk0Ma5rfO6QSaxDPzo5oiWje4NjZAked\nblTLwjQMjqdSTOo6lxSFS0tLdIUjJBfBkjJ4PQ4tnTIf/HGUg+39qIqC9t57RHRdhMB4POJ/nw+8\nXmFs//KXjKyuMrSwQGSd+ybo8eC/+D7/5Y//kIF9V2eHrs27Mjmpcf/9Mu+959DUFKS9/Siy/Bma\nBIpFOH4czpwh6PGQsSxRnrijA8bF5EsySbqpqa4E8tQUvPDCWnWQSAR+67euHWd8773w9tvCdhoZ\nuXF86a8EbFt09sQJYQBW0N0NX/5ynSVdlxsIeKgc5tSdTBJRFEH4NDSg+3zMWBYD0ai492WsJ4Q+\nx98LPrP2yI1QS7T6fEJBGO0/wmsLp/l68ByPD+3hxIlZEgmZpXyYl3xD6J3/jPuz36FkXMFXyuK6\nklgM9fdX9fm3RLZJsni5zjWVFnCN3FiuC3/5l2KekST4+terfaoNKfF6vbS2thIbGmJ8eJiWbJYD\nHR18sLjIXDZLslgkEwwS7e4ms7hItlQia1mkLYsfZDKMpdPMlUqEFYW8aeItFlEcB58sU8xkeDmX\n48nmZp6dnyfhuqyGw6QvXaIlmeSMx8PU8jLPPPQQqs9H0OOpEhvdu3dvmJjcNM1NJS//1DE6Cs8/\nTzAYJLO6SmT7dkFelXP44PeTDoUIPvKIILIWF6tEBpoGThKi8xDM4VNVTB38jgemG3FXwuQyOjPO\nLPGmnYQ9PoLZSR5/HB58cM1BuFWQJKG2SCZvk9ICbk1p4boi/OYXvxAr7Qq2bePgt77FsZ/+VNhj\nNbtUw36/8Q1OLi/X1293HEG2lx1GjqoKxUwlKVctLEuEmVgW2bklpLjGYkEINgCikkyj4sVSvEzZ\nyzR7T/NYZIyOlgaRldz1gaES8XqvCqm6toPoPEf/2e8gpT5g/KXn2KXn1tQ3gBaXODVT4MHf/5eo\nvqBYgL/00qZUTOtJi5vBRhVGRvM7aM/F2bN/N4qsIAfaoSEMloauZVF99XVaZYTDzrZNFkfPYKZT\nSI7DzOgs++6TcdwosiTGrZGLSYa6NSINQSYzM2C2geQhEvQw1LtSTfBfi83Y9cAnsv3/e8RmSIsB\n13VrDYLvSJL00e3q0FbitoaH2Le3+sCOHTtIJBLMzs5enddi1y6h9QahtrgGaTE2BufP59C0VTo7\n4d57t7iUvevC/E/pDE2QA5a03cS9X6X1dhAWILIHplIc6e/nQixG0rLw1dyXihy1Yjw8+bv/e5W5\nfH9xnnSuSEmKcsdD/UiSxML0BYLhSXweQUr4Q62k5xIYhkKh6PD6RCOtwSIDXTp+j01AdQh6Tbxa\nCblswK3GZSQpg9zlUspLWJZMKpvF45mm1a/RHlS4rzfHXXtTyDJIkoRt24z+9Dw7B3aCrCApXiTF\nz7sfjvNg0zSNlh837UGSFUAmIkkMtYwz8uIfc+jBB8uGpPgMSUGVFL7yDx/m4rvvoy8sIJkOtkel\n5Z8/ym8Q44MPf8iZ5Rk8mY/p7Amyd08HilICVwa/QlqXCX3xIOZde/mzP/8ucV3D19sDSMQXxjiy\nw8OlpgYO3Xsviu3gShKRfJ6hmMnIWZWDwTaevXiRRKkkMsErCkuGwekrwzyx3+VwSy+xqExTE2S0\nLMcurXC055BQdzjOmkQSMB2njvyQFYUnZJkj0SiqzwexGObgIMfm3mVoX5H+5mtnh65VowAcO2by\n/e8f5+c/f45Ll0z27lXp6ekh1gBGdhbZNTEdmUtTKfBGcRwHVb26zbVY8vWG7Gb2uy4qhtGrr1Y9\nZwd37+ZYJMJQuKxQmZoC2yazssKpgQGOHjmC48Bbb4l1fMV5s28ffO1r16+rHo0KYcKVK/DRR5Qr\nDG3ql/npw3Hg7FmR0bbW0mlrE2TFrl03JE7XV/QZ6+6meWqK5q4uBpqbUTKZKmmR1vU6Quhz/L3h\nM2uP3AjriVbTlFFVhy/8i8do4G7emp3lVEuCdEMrxUgf2vIRHEdlxjvAYnMj4dwCHdl3xW9D06pe\n3lsm22QV7NI1lRbXxOnTMDsr/j50qOaM1usAACAASURBVC4300YhJYqi0H3PPWiJBD0HDnD2xAki\nmkYsEqGxvZ18Ps+MLKPOz2N4PMiKAp2dpGwbyecjbZpogQCOpmGYZtXwG9U0fpJIsFAqEfJ4SBeL\neBIJFNvGK8vklpf5SSLB01/6Et8dGyNRLLKazxMqu3zn5ua4cOECzzwjFhfPPvssiUQCf83Cbnp6\nmhdeeIEDBw5U541PQnasbyPLMqlUimg0es3jAkJl9sor4LocHBjgWDQq5oiaMbC6aD5yhJHjx8nM\nzYmcFjEvBDPg04FmdLMRpb2L1IQEq140zSCxlMIuGvg8CWjaSUsLPHrQ3JL8aNdCJCJIi61XWlRK\nk5REBZSbCfN2XeEoeO01kSSqglhMzDv79qFKUt28slHYb7CvT9z/ispUloX8IRgUc02FWFpdFSSF\nplWTPlh5ndyqznhSxU7l2LXuZ6kqDk0NOg0NOneEDdzlJTo0DeJXX05EVdHGx2F5mZHLlxkaHycS\nDIqJX1XBC5HGNI+Z77L88ggoCjsqhIXjhVIH6B0EnSAH0iVGXn/zptWxDQ1rqpqbJS3g6gojD5VJ\ngm4tSSSk4CADMrrjZ6YUZWD/IZBdcAxwDJxYCKvtCKd/9H3uzOs0+AIgG0wZFsFMkZRWINrVjSzJ\naFqRiN8CM4tU1CD5wdq9BLTRi9ByVjxTih9kHyMjlxiKXSFCAPKKyKsiqSKfXFeGkTd/DLKPoe1x\nIuF6Ay0S8tVVO/wcmyMtipIkPey67kkASZKGgC1ZsUuS9ATwLKAA/8l13T/eoM2fAU+Wz/k/u677\nwfo210JlPes4W1TsQ/l0lBYgSIuRkRFM02R+fp7e3pqwjuZmscJIpQQz8eCDV+1v22IOS6cnURTh\ncOzbaqN76TVInaWxEbJmNxdzv8XOBZnWtq09DSBGtLfeAkCNxXjmj/6I42+8cV0DoHYw0319nDx5\nknBNKEJn7z5OTWV5dDDHnt0DDA4OYlkWi8spfvmRh1137Ca1MkHeWsTAIamLJJ6u6+LYJqZpcWYp\nSHtQYzZuokgOfo+NV7WxbB1ZcsQzZ8t18xtAIpGoymMr+Hh8nu47TLK16n8JJPEPZ89/hL38FpIk\nIZUf5srflZfsk5D8EookIWXP0SxJPLpDouuLDlZqnLTP5JKtINliX9NxSPi9dOYXOPUXf0ZvbpUd\nfhlJknCRmHKTDDbaOCbMXZwlFmvG15gh4+ZQwx5WV0ze9eaJ7S/SKqkYpkugrZN4XuIrd0fweWGh\nuYG+L3SBbROxHYYGdUZWFDx6DyRSYNlg2pjFEt+7Mk5c0/FJCgVZxS2ZnDRMLhQKPNPdjbq8zIhc\nZOjuIpGAh5SzllDzegO8aZqcOfMsq6sJDMPPhQugqjYX33merx+Exx8V1tfw8DC9pRxvnA/T2CMI\nj9o2GyVihasNWdu2r9rPdmw+PPcW/+pHz9Hc/zB+v//apInpxTl3DjWdpicSEW1avch37Ma7bw9/\nO5WiyxvF29WFE4/jbWrC3r2bP/3T5zh92iSbVYlG+9i16whf/arKwYNgWSbDb11ffnjXXSYvv3yc\nVGqSRMKkr++TkzYbGeTr5Y/exh6SBZidna0a6etzw9S1MQzklRUYP8vekI3qAceV8TZsI9l/F7O2\njfnyy8ivvrphjpmNwqwq6ouHy56mHckkyvS0sJaLxTpj/1rX+knJr5u+/k/Q5tcMn1l7ZDNYT7Su\nh943zMmTewn7IizqwgRYkfvIWXNIHj9+f9noyeUgFNoask3yAKXrKi2uQi4nFnYgVp+P1udJ3Uy5\ndWvnTk6ePFklNZqamujs7GR8eJj2fJ6He3o4uG0b2VKJuWyWWUmiubOT1akp4Z12XXAcHEliyXXR\nXRfdcQTJUSqh18wdE8UiP3nxReZlmYDHQyadJh8I4PF48Hg8zMzM8Od//uf4fD7GxsZoaGjAtu1y\nTiWZDz74gFwuh67rDJbzjG2G7LheG9u2GR4eJpfLEQ6Xw04VpZ4gsSzk0VFSly8T9flwZBn14EF6\njh7lHNRVk/H29GADzz33HLqu88Mr5/jNzjQPDigosrAnCobFT5cbyKj38eHIu7TP5QiWojzm7aSt\nOE2AInt3FhgpLvDKhxaXn2XTBE0lzLl2jLoeIVPJa7FeaXHLShdvjf7BzNaRFtcljZJJ1PFxesrS\n7dlMBtPjQR4cJAVEX38d5/jx696P5557rnrcV1dW+IPmZlqCa+uKuKbxvUSC6JUrTJw/z5mpKfa1\ntnK4p59MoZGVFMSTNidWP+aNYAxZsbjoWPQHQzy9vY32sElQMbBKOsfn5nhlJYNTKnHJNOnz+4Xz\np6x2Mh2H40tLvDo3x9lslonZ2bKDqAk1YkBLHttXYCKukbIsVqdlJElmW9rPfqsbtdQAShZTynE8\nHmeyWOTCyZM8/xd/wQO5HMH+fqHsKJ8vIkkMZTKM/OAHHHrsMbEgkyQkWWabJJHMSxSnJFiRKHv4\nrn5ttH3dNlWWOfqbf8jI8Oto45OM5bppnp+iubmLgX39VRvOliQ+uDjF62db+Nmb36d7aopkaytH\n+oUS2NFsPKUCDZrJUriVbb3tyN4cyDlwDFzEb8ZxHVKpFLquMzOl8N6wS7QpSv+AOJe2cp5IVLtq\nlnJcByuV4uLpt3Bdh547LUx/kFhLO7LHD6EBUCNEQj608bUQv80Qm5v5rX2SNtfadjPk661iM6TF\n/wL8F0mSKr/0FPCPbvXEkiQpwHeBx4F54H1Jkn7muu6lmjZPATtd1x2UJOkB4N8jsodvCrVVQS1r\nC7yGsld4uF376qQzW4z1eS3qSAtJEiEi770nSlqV49lr8e67InIklZpkxw4Ihbx0bVkKZiA+LGp1\nA8FYM5f0b+GgMjcHd9+9daep4vTptdnr8GHUQOCmSrVey7Pja7uHt5cTNOw+wPwVh7OL27kwHqJ9\nl4hz7+zs5MzsMI/vLRLwQm9vL/F4nMXlFH/3gULa3UNm8QL7t2m4ro3j2AQCMl6vhYyFLLmMzCvY\nqx2oiiNesktDwMPdTf0okoVHtlGw0e0URbuELLlIOIL0cAFcJFxc18YwLCTJve61bgTHASsQ44qd\nImLqSK6LK0kofj+hSITllWXm5+cx10kRbaOAoQvjLmPpuJaO67qkjSx+x2FZV8g6oAQcHNfFCsi0\neArk8jl27HRAgqA8y9x4kyBgJAkJiUsTF+hpCfKuksKnKEjASqFAR5tGlyxjOw7bAwFcw6DBtLBs\nh7fdFF2qn0nNYjCgknAlXFMleSWDOLiYVGc+/pB2+zSSrCDJHpA9XLr0MeHMFF+928v0tIztyuSu\nrPCth5I0SBLTF0+AJKOU4myLePj6fp3Xxs/gIvHNAzkCisT0lQv09w8AEpGAxFDPAiMnfkZal8im\nVmjw+3FwcJFILU/wxJ0mCjAxPkH/QD/Dw8MUi0Ue7IVXL58h2t5/NWnyzjv0xhd54x2dRlf8Xi/m\nL/P1IzEef/oJFMUAkmS2lXh7qsSjv//78POf8+zwMGM/PM7EcnM1HM6y5ti58wIHDjyDZd1Yfgjw\n8svPsrCQwHX9jI5CIHA1+bLRfpsxyNef33ZsXnvjeV4ZgeY+cf3Z2WEe25Wr5oYBRJvTLs1N+2Fq\nimz2Eo/daRBs8HJoYAB6u3nt43FeeW/8msdRZOWGUss65YVhICeTOLJM8P77Ofr009XkfTciqGBz\n5NdNXf8ttPk1w2fWHtkK1IaQhMMRUinQA0d4STrNN9UxosGIWKxns2RisTqy7ROjEkp3M0qLl19e\n06s//fRV9gncuNz6jdQYoQMHmHQcrMVFGj0eHu/vR1EUZlpaaCzn8dAtC7/HQ75UIqnr5AwDNRTC\n1DTRv7Is33EcUpoGDQ0ULYtCqURiXS6BVCoFUK1sVkEymSSXy6GqKrquY1nCuaEoClNTU/zJn/wJ\niqJw8eJFfD6fyJ9TDjeVZZmZmRmef/55FEVhcXGRQCCAbduMj49TLBbx+/0Ui0UmJibo7+9fI0g0\njX5dZ/ijj8gZBuFAgENf/SqKojD3zjvEYjGeqUk6Whm3wgGF3sAMBx7Ks5pM8dKKzb7e7eTcbv7b\n+RJuYyfZ8xpnp/3sSCwjk2VOlvlXEWgMw5+dO8GCJKENDJAuu8ZvRNDYts3zzz8PwNDQ2hi5npCp\nPU4kIp67XE444RSFDcff9ee/IXFRW/bUyIBfeNnWH7tKGqVShA2DQ+U4hudnZkCSGHr4YejuZvj9\n9697HRvdDwCtuZnvJBJ8vbsbr+NgyDI/m50lEIvhJpMEmtt56+wkH01M8LPhFX6rSww1P5h5m7NA\nrGsPajiLVFhmwavxQinHM/ccwgKeHR5mwXXR9u3DTKVYzeeZK5W44Lo8s3s3mCbPfvQRCdsm7feT\n9vkwPCaz0Rwnti1wuEOotIazWYqOg6krfLTsY2bJRzRnEfMs8Uy3WGA9Oze3VuHH46GwvMyZUomp\niQnhaKopIxMBtKkpoRCtwX1nhaom9DGweP2vbzNQKQ/MksTDTifHTkyxY+8syqzwHtquy2tjU/z0\njESzuoPU/DxNpsnJsTEunD7NM319BCMKGc0mEvRSyhTB34bja4dAGL1kosbasSOdnP3wDEZJQfWE\nWCh4UZNRnJUiHy+O8+RXvoysXgClHF7nWuBSTvg/h2VaOJa3/PxZlEopNC1Hd1c3sjQt8oMAsmte\n/xmtef6AG/7WPkkbRVGu+h1XIgBuhny9VWymesiHwF2SJDWW329VhNn9wBXXdacAJEn6G+AbQE21\nZ74OfL983nclSWqSJKnddd0NCvdcjdqICtveAtJCkkSIiJm77eEh4XCYlpYW4vE4U1NTPPLII/UN\nKqSFbcPEBOzZU/0onxeiBNd1MYwJdu2C7du314eY3ArSF2DxVfG3GkLu/zbt24JMTa0lo95SGMZa\nhsDmZjhw4KYPsRnPDsBDXxODQjKZRFEUFEWhsecQL358mVjQ5PHug3y0eIUT7yVIa7sBBUcxQZmm\nLeYhHLbYObCdZEoYMpm8ga+lnXBzL67rVl/BpiYOHPltHMcRyg3HIXfqxzR0pUQb3Lr2jutiNDdS\n3Pk0jm3jOBauY+E6Nq5jiQRDlffu2me4Nq5jM+7YzC3aKLLFQw/ZKJIjPnNt8o74/8NlC9MolQkT\nkbSolDdYLdhIuHg8CnKwAQkXqcFPRiuwZNr4HQckG6/PQ7DBT7EIanmeUj2A62Kvi9GyTRfZH2PO\ntomUSnhkmZVCAdNxcCwLXZYJKAoGYnEryxILpoHkuBSsEjnDh+O6mJKPbLJ+lssue5i+UKzblpif\np6FMyAw2C3s17mRpxMFjQX55BQC/7YAJzRKE8jMAtEoOGFCYnWA6+0GVeEGCC+dfIRyQ2OOYoFNV\nwASMNNu9DiBhLIwzvnqKoKYRlGTwSnSUprFn3+N39uYJOhJj7y+BaaGurNLlc/nGA/D62RxOoIH/\n4V6HhsAqo2depKVFGFiSJLNXtvjpqUtIs+fIOymC8QID4XKZzohEW5uEnbd48S8XAZl7Wi7j5FTS\nOaGiEV2VuCNg8dJfp3FcGXPlIvfs8LKyAtgSqSuLPH3HCn4HRs+dorurGwBZkjgQnub4j/9fTFum\nuHKWiOoFg7JKB3ySRHF1gZd+/D38qsS9sYt4XS/FrDjvzMwMPjfD0/tdXr1yBpA4ujtBOKhiaCnG\nL58FJLylJF/tyHJseAnXha8cLBEKeCiFAkxEw7iFNKpU5Kt3wqtXPhbH2ZsnHPBQKhaYuDLGzp0D\nNAZlvtizwvsnfs6hR56ofle1EjyPDIeeOAJ9vfDDH4mNX7gXFAkcm+PHXiGVjBPw+wDx/aaWx3ny\nThMPMFHOMVIsFmlq9PP43iKvjoucQrUk1uDgIBPjEyiYPHkn1TZf2VskHCwvUsrbbrXNrxM+y/bI\nVqA2hOSXv9RYWZGRZYf7fvMxGlrv5s0f/Qg5HscBgg8/fN0qVJtGpYLIZpUWly/DxYvi7/37N1Xh\nbCNsds7Wdu3i5MmTVfuma+dOxuNxeotF/B4Pj2zfzmQqxWw2y0wgwKFDh5gbGyMyNYUsSRiZDA25\nHAXHIRUMUnBdnGiUUCiEaZpYloVt2zg1yoxaFIvFsq1l4DgOy+sSJWfK8Q3ryY5ajJfHjdo28/Pz\nVQJEkiTS6TTj4+Mkk0k8ssylU6eYkyTimobi9VJQVYYvXKCnp0eEvy4s8N3vfpehoSHeeecdRi+d\nY09rmgHPMh7bpqRLeINBFnIhUtLjlOQmloqnmb9SIp83cNUYl5hml6eEGXF52SgixW3GSiVWe3rY\n0dZGoVCo6+8LL7xQPbfP50PXdSRJYnx8vPr32NhY9Vp9Ph+apjE+Ps7g4CBer5dEIsHx48dpaxOE\nlusK4qKpCY4fP04ymaxb/AP4fD6SySTHjx+/sTNLrVFaWGvDx/pjT1y+THF+Hr+mUXQcJspybbOh\nARobmbBtmJ6+ilgaHBys6w+wYZ+DwSCKomDt3MnRp57ipZdeIji3SC7nY3ISEgkF23uI5cAEFFZJ\n5JeRfB6SPa08dM9uolEFx+lnfHiF3mKRZLHI8Qkx/i/k8yyHwwz097M4MYGezeIPBEhaFsfLuZ+S\ngQBSwE/3vhh37PZw1x6XmOXDclQmQiFc22VyUSW+2kA67WHS64OwQ4eikTRNjlsWuC5JScLv9YIj\nQocl18Uny+JcqRRPlUuHVyBv8Duq3JraHNpbAtcV4TpdDzAytpZM/uxynAuLftr8LSi2jWRZYNv4\nQNzH+XmOGK0cO+9l6A4DyS/6HAxHWU6nyJghBvbtZ2J8gkRBweNpYiVjsWBtxyiI8a6ULaGMdeJv\nOwwtFVW1C67N5NhlFrI6qkdituhFwSFlFpAlB7+RJ5VK0dy6psBxJDHWXfWMTkxc9fwBVedj5f1W\ntBkcHGRiYuKqbcB1n/f1fbxVXJO0kCTp267r/rUkSf8Hoi56ZbsEuK7r/ttbPHcXMFvzfg54YBNt\nuoFPRFpsCSqkxW0ODwGhtojH48zMzGBZFp5a6ciOHWuBYGNjdaTFL34hFmW6nqKrK4MsQ3//FuWz\nyE/C7N+KmUTxwY7/CbxNdHcL8nR5eUPhx62hNhv24cOfuAD4jTw7lTYbGUoPPvggX/rSEUZGVIqj\nJgXzWSwricejEG3v51xyhSe78jQ2hugf6IdxWFpJc2I8wvbBu+sIo0qy0IMHD9afvJQkUlu+qQbp\nnM4XYo984ri2ixfhR+U12BefEHnarjrH9753VciKPj9K0TdNg1/GGwjQf+BubMfGsR2SmSK72+7h\nw0szxONxHMdhadxmfs6i1XuJge1FmhodfJJKQ6SvTKg4uK5DIRik0HIIqVHn0tgoRirBpZkr2I6N\n4lcJRyLIJaEuya0s0ehYBB0PiqqQtG0yeQVN9RAON2PZCCIFQHLJWV6yZiNSWbEi4ZLVPdi2i4yL\n7HExiyADhgmyIkIc10Ndx/G5rovrCEKpus02MQyw1mX7lVwLx65UZLGxrRJSzWTtRca0CkR8Djgu\nRjIHloXXI44d80oEYxLIFi0BsZ+ZzVJwU9VjyMDy5BQ+N01nh0mbnMbyhQgEyr+/8pinTY6S04E7\nTDJru9dh8fx5wn7oxmRbE2TLP7FUPkunzwEXzPgiidJE3X4z598n7Idex4TSxsfOn/0lizrsvsOs\nG7hLuSxNjkOTCm2laQC6VEEaAehzV8C0iFoW0Si0dQpyoaurXLGGDMUFQVjFbAcUaC+OAtCtCKIJ\noDj7MdPpNe/a2fM/wbj47MadrcAwILcg/v7rl6r5AVLz8+wyzeqxATxaVpzPhuLMRSQkWhwbDGjy\nQKd2CSToqbb5mOnMOxQzWZptGxTo0D4W16Y4SCUx4Wozl8vXVmlzGRfoUmyh1keiMDNabeMq0KaJ\n93uiDdXkYb8O+HWwR7YKlRCShx6CP/kTsW3fvnKUqN8vSv7Ksqg/6dmMmPYGuBmlRakEL70k/vb7\n4cknb+nUm5mz1ysyFEVh4NAhZi5fxm8YeO+9l+z4OIvxOAPlSkHdg4OMJxL0Fot4w2EeURQmdZ1Z\nr5eZtjbuLHvMK3Bdl6amJkzTZGlpCcuysCwLwzDI5/MUi8Vq7jG/34/jONi2fV2yoxYbtak4LCqE\nv67rlEolTF3HyOcxJIkcYFYqeGka+tRUnVpydXWVXCZFauI49zatELBs0knx2VLOz5n5JlYKLv6R\nN4jHIR5fs2sVBaJtMmMlHSev86GmgWFgKgohIPXB1ZFRo6Ni/FlP0NSqOCs2Rm0/V1ZWWFhYqL4/\ne/YsAwPvce6ceJ/NSkQi8N57712X/Dl79iwjIyNrhPQ6CIeDw1DTScBlRp9lpjQACE9zoVAQ6ptM\nhtmZGayahUPctiEYxLBtKBaJp1JIkoRhrE0GiUSCeHwtecTo6CiSJJGvTRK9Dpcvj3HixBhvvHGS\nZLJQW0QEEN9DKKqQb0vS0CChFvKMjZ2ufu6Wn11b03gnlwNZpqQoRBSF9MgIjuPwYTxO1DRRZZnx\nVIqQ36YllqWzy6I12khpVULXChSKRbKayhvnFWZWAyQLEqBhOg75chWtxUIBVZaZKCuf85VwE0ki\n39SEo2msAjgOE4bB9Dp7/bLHw6VKaXEA12XG5zLhgmS6fGC4eBSxvfoqt6v+v9H22gp4G7XzN4Mr\nruFUSiPvSsyVF9EJw2C1YsM5DhPLy0yXStiZMMOzMolwkW0fn6dkSaTjEt98ADJnzjA1NYVhGBRK\nDq+cUymqYaYW1r6byclJBrc3I81fIdywRh5X9ssXbYYvC1KnM5Cjwa/Q3mBgl/LopRKL0wrZgsUb\nszsZXf4PvPnmm3XPUuU4FVSUT5Vt69/fSptMJlN3vsq2WkyVFTS5Svb3Dfp4q7jejFahecLUGAlb\niM0ec/3os+F+3/nOd6p/Hz58mMOHD18VHrIlqOS1uM1KCxCkxenTp7Esi/n5ebZv3772oaqKRBVj\nY3WlTxcWRIlugHB4opp4b0vyWRSXYPpvRHiMpMD234VAByCS9YPoxsKC4FS2BLouyhiCSLJ3xx1b\ndOBrY72hZFmiQMG/+3cV7kTl0KFnyGaP09w8SWOjiSx/jWkjRbAlyskJB9Nt5eSSH19bc53xU5ss\ndD02Kt8EkMmXODXdzNHf/uQy39rvY2pqY9KiNqN7BbGOAV7/eJVHB3P09vRWq7Rk8iUuZrfxzd/9\nPRqOH+fkyZNYlo/xcWiMCHVEODaN3w/be7czUONtS+d0Hth2NQHzvQ1IE4CGPpvFiQnMTIYOyyJt\nxvHEotz/8KFqjpHKK5nTuXv7F7nvi0dwHKf6OvcXf1E1JlzXJetxmFv4gJ5tRVTVoafbhyS56KVC\neZJzmCx6kXBYNnQkycXnVWlq2i5IizL5YjUGmc45aIV8lSBxXQdNmydlGICL6lHKeVAsKqE+cU1B\nVWzyRQOMEh4JRB5rF0n1IHk8WMhILhi2KxLAI+G4MpUhsGRANucSslVkx0SRLPwNLoqnfsh0XRfP\nxjZcFR7Jrc7tFRvYsgSZ41IehDcYeWv3uxauef6a/eoIItcRJ3fqD6x6lXJp3Y2PcdVxrtHGs5nw\nqmtMHu4GF1vfd8rPR+3n7tqHUFZFOeCK5wXAqwiDq5LJXAIq3axs8yr17wHkan/ENr/iMr5k8t2/\nK13TcP+M4jNvj2w1GhsFl5bPizkXgG3bxP+OIzwIWxESKpd/C5shLd54Yy1z4pEjV5V8vB24nqOh\nosj4knm1grKO2LBtsouLLJa3r1ellkoldu/eDYgwkcbGNRJ0YWEBTdOwLIvt27dXvY8gxouWlhZs\n22Z5eblKZtTOTY7jEIlEcByHRCJR3VYp4V5RY6qqiq1pyMUiTjl23/b5kL1eXNZIjgokXHZE0tzf\n9C7xzlUsSyzAk5qXkfko81k/ritRKMDiolPn2GtsFMoGWZYhJBZVfsOAlRV0yxI5Q3xXO1euRdDU\n9mujMXT9Nsdx6g5fLLo0Noowk432r6BCFN0IRUvBJxt4nEJ1QWXoOnYqJbJBVr4b1xVzgd+P5fdD\nWRULa46K2vOZpkmptMbeV8ZgfZ2EwHVFXs1CQfDjyWSKRCKPZYl2lbycDQ2C+5MkcY/y+auPBeAL\nhSAUWlNz6Hpdu2BbG5lMkt5wlv1deTqbTFxZOKIsy8K0JcbTUd4/66GUcgl5FcBAtywsxyHj8RAt\nLyYWXZdIqYRd/h7Wt8nrOhnDwCPL4rmveSZ0yyLf2Vm3DSATkFguKz+njds/bGSCQfQaMsWIRMhk\nMnjKZIpdLLJsGJBIYCptGF17sMKDKEBT4x7+v/MT+NwU8zM2mg7pUhC8UWSrXklgWRYMDvKT0w6P\n700TDoqxVNd1MnmDl896KEriO/vpaZcn79IJKQ5WUFTkiyfjvHyxkcaeGEtLS6TT6brvtRKOVsH6\nZ3P9+1tpk8/n685X2XbV9VL/jKZSqTpF1q3imqSF67r/sfzna67rvl37WTn51a1iHqgtvteD8Fxc\nr013edtVqDUSKrhtSgv41JQWFUxNTdWTFiBkl2NjItfDygpuW3slgTQisfYEs7PQ0NBA+80UP94I\nRhqmfiAyLgP0fBNCayvfWttobm4LSYt33lkrO1hJ3vMpwXFEFYU336zPYN3RAY89pjI4+NR1u3P4\nGzeXMGqj8k2VBIJHf/vWZL7BoKh/vbwsquVuhM3m/VjfpyNHjnD27AXeeCOJ44j99t3dz7ELC3z9\nIEJ5Usb1CJiNSJNKH1p6enjkW9/i8PnzPJTPc2xhiu6iRSTkqxoGmXyJ9+fbOfrbX7vqXt111111\nSd3CYUgud2HY06IaXGA7ba0wPTONx+MhV7BQmsTvzVCmCXhhW+92eteRL3e3P0KyQN2xAQzGyJf3\n6+7dTiqdqnqIsgULraED38oYVlo8WF5VhWAQPRQCVcV1XTIBMdbk/YXydxik9+B9mJbL2BisFlyW\n5GZmrizTnvklkQgQ3AfRKK7rPuQhLQAAIABJREFUlFdXLrFolD2d0DqYWFOJuE6ZbCgrO+QoVxZt\n0ukUruuQMmFi3KWB89yrFQkGXQK+AL39+6r7uK5LoxFhJg6pdArKahdhVLrV1V1TMEJXl0uoJw0I\nkkMC5kcvUyrpSMCi5UWybVLZFOhFcH34PApIYHg8EAyyLPnBhrRTZu0lF58iJF2GLeQJK6Z4n3HF\nOCUBXsVLa/Pa2Km2honsXisjXTGCr1qNLrwuSFNfF+wSsaWJ5ZMUSvUTddqcJlvuk9fjBVwMy6ge\nL2F6cXHJlRd9Xo9Kc1M3Wm4Os7xfwhTPa95Z80ioHlUoXMr7JUxPuc3a4lFVxH5WeVvK9BBrhn/w\ncIBybj3+9O+2Ov3+p49fB3tkqyFJgqMYHd2AtACxcStIi7I0+YbhIQsLIpkWQG+vqKH8KeFGiowb\nEhsvvsiXzpzh2dOnSZrmdR0N6+fIaDRKKpUiFApdpWgtlUoMDAhP/vLyMoENyjfpus5DDz0E1M8j\npVKJ6WkxH1mWxfZwmNT4OFoohOU4bN+9m5THg1a2j1zXJRgMct99B2lR4/Q1TNIa9nPw3oOcHnGJ\nZ12u5HtYUprp2iPRVHAZH3cxTejqigEOqppkxw6XQGBtjK/8Hw0G4Z13SBSL0NGBW37Wats0lfM+\nVPJ/VIgU0zSr0vDK9em6Xt3X7/fX2buRSISjRw+zuio+37cP9u93SSaTZDKZaxIXkUikGqNf27f1\n2FZM47cTNCsdxHwPwtQUq5pGxrKqK+YCUPL5cMuy4drvZf11VODz+ejo6Ki+r5Bb2WwWxxGkQyaz\nlqcDIBxupL39LvL5eXy+LJEINDS4V9mWtce6Fta3cV2XJp9OX1OK7REJVQ7VXUeiGGQi3cRcNoLl\nyjTtdEgtLyPZtgj3MAy8DQ3sam+v2lktLS2kV1bAtpEcB6fcZrCtDVmWaWlpYXlsjIhp0hwIsLdc\n3VCzLD5uaODRp56qV40j/JErIkKXrq76Yex2YGZmpu4+xmIx0WfDQFUUGsNh9jgOmm3zsSxz+Mkn\n6/u8fz8AqWPHsLJZrlV7oLGxkf3792Pt3s2FiXPIuVU8ksX5KyVW8jLe1nbayuSJ47Tw8pVlOgNp\n7upME41FSXrv4guPf6l67tp+u65LOp2uI8mu9YxuRZvm5ua681W21aKSs6JWaZFOp+tI3vXqjJvF\nZrSDfw7cs27bnwG3OiOdBgYlSdoBLAC/C/yP69r8DPhfgb+RJOkQkL6Z+NHbQlp8ikqLUChUl9fi\nS1/6Un2D2ljRsTHOr7QzI8Lwuf9+l48+mgLEYvCWvG6WBpP/VYTFAHR+BZrurGsSDosk4ZnMFua1\n0DRBWoAYycqejtsN1xXhFL/4hUhmWkEsJniT/fs3x51sRtq60T63q7RRX58gLebnNw7h2WwM8UZ9\nvuOOZzh3TlSd6Osz2b1bpafnX9DQAG9Nzm6KgNmINIEao/HoUdA01AsXOBrcx4hxF9r41Cc+9p47\n+zn2ywWeuBNMs5/2DlhZXSGezPOL0RDRHmGAvnLuxuTLVYZse3/dfhPjE2SzWYoll9eGDaLpOCnd\nIac7BPwy0cFBaIqSnZnGA+Q1G0MRUsaCniPgheZYMwXNw6VL5ZLuuo63cS9f+4c7Gfuv74jyv5pW\nN+Prus6dd91DrAEsrh16tG3gAfwda0ZzWxAm5iCRzbGUmqYvDK0d22lq6arbr3f/I4Q2IG1qz7/v\nC48QawB1XehTqiBIoqLuYGb9EE+Qb4awvwHLcWjt2g5t7UwvL1EsQalBEA85d5pw8P9n792j27qv\ne8/PwYMg+AKfokhRIkGKelsPU7Jlm7blB+XUzqtp3aRtepveNje90861c287t3c6f2TWzNx1p6sz\n9Wqblb6S3nTqNLdu0sRpbUeMLCmyJcU2ZVsyrTcpipQokeIDfICPA+DMHxsgABIkQRIgQWp/1tIS\nCRwcHBwenN/+fX97f7cE8huq5LGOGx2MTcB4nvw+FJJtzECAiqpqKurqp455a/Fj7EvmO3YrT4yO\nSzbCM78JwO1g/YzP6veX4LNEoKrYGD2eiPg1Gj6mwfA26zdWs6G+Hr9xZeq4I9sMxHy28o3xny3h\nNpvitxkOb3NjKLrNGmPVxiPpICJa9PXJPSG7qEj6Go+NyY3+wIGlv0ky5SGhELz6anTF5FOfWtYF\nhmSYczyurcX54Ye8uH8/zZs30z4yMuv4N32MfPDBB8nOzqakZO6syjnHtgTb1NbW0tPTw8jwMHmm\nSa3NRpvLxcDYGHlVVdQePEhbWxtDYWEjGAyydUMWB4o/pMAxTCAQoLSkmqzcEkIVz3Hm+i2yXG4i\nnbK7uuTPZbONU1u7l337oLPzJNnZie/jDz72GAwPc/LDD3EFg1BWNmObRx8VMXj6PXJ0dJSODikB\njCy+xQky1dVTokWkfPbw4UO0tEjFUV2dxF7j4+NzjjWPPfYYTz/99KzXwBQd/eC7AMMW9310Hbq7\nCVRXc7KjA1d+Pni9WIODdNy4EXeMkeOe63NEMm3Gx8dpbHyMO3fgtddO4vO5phoChCstCIXGefTR\nx/i1X3uWK1eyOXVq7s+W6Nwm2ubUW8fYlO+jIrubfMcwMtUrIRAIUFW9hSG7lyPv3iWQV0RuHsRG\n1uP19Qt6r0TbBHfupOPSJepLSynfsmWq5esXZ4kjR0aiosXDD8OD0wvzUkx+fv6M4w7u3El3Wxv+\n3l621NSwzukkZ3CQX6utxbl3b8Is74KCgnnPUaJ7zmuvvTbr60KTQ3yi4Y5cR5WfgNKo1/P098vO\nzp5x/cH81+hitqmvr596v9jHpn9eYM5jXCpzeVo8BDwMlBmG8R+JLkTlIy3BloRlWQHDMH4X+HF4\nf9+0LOuCYRhfCT//l5ZlvWYYxrOGYVxFhM/fWMh7xApjKc+0CI5LGnOa64a9Xi93796ls7Nzpq9F\nURGUlsLduwQuXKF5WBTm3FzYsuU2p0/7p/axaEImdPwDTITr9EofgrKZLVZBSkQiokVKWsy+/bbM\nrmFZsiwsC65eFbGiO8bbsaBArDT27IkXwlYbNTVw5ozElzduwObNM7dZjNBy9Sq8955knlRVwW/8\nxuLOU1KiSV0dtLbinJjgYP1e2PDckvZd92tfpeUUXD3XyakbJg0PfIqOkQHq7o+2fTpwYH7xJdG+\nY19nhQo5ev4u/Vd6WWfkYbdBUVYRb1yDTz9dTG3Y32Q20eRT94PDWcsHH4SNyfwTfNBXwh/+703k\n5sJLR1+m/9YtXP39U595ekCcTOlRbNBcUQHt47X88zu3+O3yhYk2ybx/bXU1bR9+zBvHfBQbVeD0\n8JNzwzy530ZpTSW1YZG27U4fb3wExV55/59c6OHJLSOUFudNHVPbjVvzbrPgMqviYhEtYs5pIvFr\nukA1198xdpvautqkjnup26wF1kI8kg5iVyS7u8HrDadfXLsWk36xRKbKQ+YIOM+cYaqvd2PjjAlt\nxhOOkZx2O896vTDd+DyGRGNkMm04k1kQmL7Np559loHTpykaGiIElBUWku12UxI2Vo8IG26rn0e2\njXNgu3QkCQQCOLML8D78FSh/iAc3w6mLL3HlSj83brimGrsEgxNs3VrCn/xJE9nZ8NJL89zHLYvW\ny5fpHxjAFQhMBdnzCTS1tbVTnhWRbJSenh5GRkbiMlSm76egAHp7o43j5l3YSLZTzlAAzn0I/T7o\nfxQwaLrvPlpLSuh3u3G53dSWlNDT2zvjGOf7HJYFPT0TjI+XcP58E34/DA21Mj7ej8PhwmYTP/nC\nwgnq6kr4T/+pCQltmrh4cWHCVvw2xTQ9shVj4H3yO84zMT4yNV+wLOgdy2ckawuPP/d/gs3Bicsv\nMbro95p7m0AgwH0HDiTXzQWZs0RKUsN2Cmlltqzisk2bKN67V447GIQ/+zOp4Wluhi1bZqz0LfZ6\nnOt1xcXrqa0vAoIw0T/n66aEzQVeo4vdJtH3ONHnnesYl4oxW/qUYRiPA08AXwH+IuapYeBHlmVd\nWfK7pwjDMKxEn+PKFXj5Zfn5t34r6ruwJO7+DG69Lj/v+H1w5KZgp7PT2trKK6+8AsCXvvSluBQ6\nAH78Yzh9mvYOGy9X/j4Bp5vPfAbGxk5x5MgRAF544YXF9ce1QtDxjzAkJnEU3gcbPzereHD6tBwO\nwFe/KpkXi2Z4GP70T6V2sroavvSltIoWHR1w9ChTmSogJRWPPiqLVanwM1tpxsbgj/5IBrDGRkhm\nQWI+RkfhG98QpTwrC377t6OrCGlhaAj+37Dn3pNPzhlcJktzc9Q25emn5dyklPFx+MlPMN95h+Zr\n12gfGMB0u3Hu2cPGvXspzoXJIRFEzJCNC9cHICsqmpSWbuTiORjq6cRpmARwsrPBy6/+RhNZWeG+\n2G++SfM3vyn7fvRRnAUFCXtnt5xqxt8fX3rU8HDi/tpDQyZvveWkoGAjD+yFPVs6k3rdXL27W041\n4797FVvHNULXOsiacNM/Gu557/Fg27oV8gy21xThtEkpUlbBRtkmpp84kwML3mb6Mc/LyZNyUwD4\nL/9lqoY70WfduHHjvH/H6duk8rPNtc3h51/AsqzMWvZeIGshHkkHIyPwx38sP0/du44elWvXMOS6\nXaorducPYOADyPLAtq/OfH5wEL7+dRmrS0rg3//71Tlg/vmfw927ou5/6UsrfTQyuH73u9AZ9n4t\nKYFf/VXM/Pyp+48jMMAm11Xc5nXcbreUQ9iyMEof4v5nfhdntqRr9/fDq6/KfWtgoJ1QyKS42Mmz\nz3r5/OcXcB+/ehXz29+mua2Ndq8Xs7R01nt9onskxN+jBgYGKCqK3iOn7+fv/14WRSoq4CtfIblj\nnIv+flmVuvo65ErHFkYOwcNPwEMPYRpG3L4THeNsnyMUKqK7O0Rvr5OcHC+1tU3Yw+V7hmHi9zdj\nt7dTXGzidic+5mTH0dht3M4Q920M8WC9E7spZTnBUJC2a23cGRjn9kQFA0Y1G2p2zvs3SubvuJht\nkiHy9du+HT7/+aRftmiSOu6zZyWDDODxx+GJJxa3n4W+//VvwVg35G8G7xfnfN1CrtGlbjPbY/Nd\nE7H7fuGFpcUjc3lanABOGIbxt5ZldSz2DVaStHpagJRNpFm0mO5rMUO0qK9n/NhpOjtCFLnbcO7d\nyd698PLL4vJfWFi4SMHCgluvRQWLPC9UfWZO4SBWFOrqWqJocfKkBEGQ1iyL7m6J8a5ejT7mckmK\n2sGDCb2mVi1utwz+t27N7muxECxL7ucRL55nn02zYAGy9FJWJssvV6+mRLR44gnp0nf3rvjIbd2a\nooXCSJ3R66/DyAhOm41nt2+XY3744VkD+1ht/soV+Od/BssJ+Rskbv3FX5S/YyzOrVt5NpKq98wz\nCevJkyk9mr6K+PLLcgzjIWh8bvb5T1Kdeex2Duavh5aLMFQEZeH7UkUFPPWUZNFkUkp57MXc3z91\n0pPNRlq8bW6qeWGlD2DJrIV4JB3k5UXLMmf4WliWZD9s2rS0N5kr08Ky4F//NTpWf+pTq1OwAHGn\nvntXRALTlDz+leLuXfjOd6JZXtXVMpPLycEJPPv0I9BjwsBdsGqAGjFHL9kPZY+CU1Y0TVME+bfe\ngkBAsiHdbvFI3bdv5u123ntbdTVOl0vGmv374ZOfTLjZYjI2ExEphY+1cVjUvkdG4Kc/hffek1TT\nLFfUFObxfwMlsmrshKT3bVkS57a2yr/hYdld5Otnt8uQtnMnbN3qJDt7/v0m2+Hu2Z/7BIy0Qf9Z\nidGtaNctDBv2oh3UH/516vNqZ80GT/q9UrBNMhQWymW/HJkWkORx790L774rE4W335YvTdi3ZUH7\nWej7u4pFtJjom/FUqs53OpnrGF94YWnxSDKjy98YhvG8ZVmDAIZhFAP/YFnWM0t652UgrZ4WsCy+\nFrm5uZSVldHb2zvVTiaOTZu40pFFKDRJSf8VHvm5nYRCwam6o0W3Ou35KfSFW/e410P1F6LByyxU\nVMg5DwblZh72qlk4g4PSrgOkhmG6AWkKiExQW1ujjzkcUkv3yCOSZbEWqamR4La7W2pFlyLKtLTI\nZB/kb71nT0oOcX42bxbRoqsrXMidPf9r5sDhgM9+Fr75Tbl2f/AD+M3fXHRnXWFwUIL5KzELwF6v\nBHnTzIsSEQiImBaxdAEZP3/u52b5m1VURGvZr11LmQleQ4N8hIkJ+OijRe7WsuSLduxYvElMaako\nRjt2ZJZYEWEW0UJZUVZtPJIuKitnES1AfC2WKlpMGXEm8LRobY3e4/btS6ED9wrg9Up79WBQUi7D\nBprLTkeHZFhEDMh374ZPfzqcPz8KPSeh713p4gZy7yzcA+WHICs6obpyRbrPDsS0ub7/fsnIWXR8\n43RKPHbtmiwapKQOeHYiC1+jozImLlgPm5iQQfTUqWipMcCWPbAxX8ZMZ/J19pYlsdNHH8mlP91T\n0GaTy2jXLti2DRL4ri6NyUEYeF8ynyanvbmrFIrvh6I9aV9MTQcRLWC5RIuksNkk6PrWt+QCPHIE\nfumX0v++WeHYwxyEUBBsq7guPcUkcwsoiwQIAJZl9RuGscRWFMtDWlqeTs+0WAZqamro7e2lq6tr\nhq/F9S4Hl8xaSrnIzqwrbKyy6LjRNdUDe1F+Fv1n4c4x+TmrEGp+Fezzz24dDumscfPmEs04T5yI\nqkwJ0rGWwuCg7D7iDQByX2pokAXwsPntmsXrlfE7FJLYaMuWxe2ntzdaClRQIHPxZZt31tVJIBIK\nScrI9u1L3mVVFTz0kJybmzdl92FD94URDEp99/Hj0dXHnBzJfti9O6mT1NcH//RPUV+VrCw5v7t3\nz/Eim01WCltboa1Nzs2SVBdhyxb5TgwPi0i1INEiYhJz9Gi03h0kEo2YxKTgGNPGdNFCyQRWbTyS\nLior4cIFGdtGRyE3YS/UJTBbpsXYmGSRgdzjkvUTyFRqaiK9JeUeuhKixblz8MMfRuOfxx+Xe2Vo\nEu6chLunIRgz+fZsh/InITuaGujzwRtvyDURYf16eO452Bjb+2ax1NWJaDE4KIpIGtMrY5oOMDS0\ngLcKBmXAOnEi0qdeqKsT1aY0Fy6Ey0zNubsZWJYYmLe2ilgRKwKBXDI1NVGhIjfVekEoAEOXYOCs\nZFfElp7ZnODZKWJFzsbMFP+TJCJajI2lZC0qdWzaJMHXuXOSOdvePuWBkzYiooVliXDhmn+h614h\nGdEiaBhGdSQlM+yunbgZc4aRnvKQmDvSMmRagAgP7777LoFAgM7OzikhIhSSmMFWXM+6/ots2TAK\n3d20tbXFvXZBDF2Gm/8iPzvcUPNFcCY/k6+qkklfd7ec8wUbMvb1SZ9RkBEgFS3bkPjt5EnJDgzG\nLFDs3i0xwWIqaFYjmzbJPDEUEgfxxYgWgQB873syJzcM+Nzn0rCiMBfV1VHXpmvXUiJaQLRMpK9P\nkgK2bFlgmUhXF/zoRxLhRNi3T4L5JJe2PvxQEjQii0KVlVIOklSwFjYpZWxMvoAp+O7YbCJUnDgR\n/V4nlXCQyCQmN1dMYvbvXx0p5C6XHPPoqIoWmcOqjUfSxfQup/X1iXqhLoFI9xArFL/q95OfRCeE\nn/jE6k9PTHX95EKwLClfOBZeLIp0YNm9U4SK3rfiF8nyvLD+KciJ1uQGgyK2nzgR1ctdLqmuPXAg\nhfpwrJhz7VpaRYvYEmOfL4m3imT1HT0ary5UVIhYETn2iIm+FQIzcQvRnp5o6cfdu/HPGYbEUjt3\nSqJgCvwFZzLeKwuIgx/OXCDNqYSi+6FwF9gzZXa/NGKrLny+DBItQK6dixclMHv9dTFvS+eCiyvm\nQp/oV9EihmQixz8EThqG8dPw748B/y59h5Q60lMeEjM7W6ZMi+qY8ojr169PCRFnz8r8yFVST7UZ\nThu/fJn28IBbXl6+MLdWfxfceEVu5DYnVP8KZJcu6FirqqRVeyAgi6sLnjcdPy4zasNISZbF+LiU\nop05Ex3IQfSQJ5+U/tD3Ei6XxLNdXYuPy44diy6cP/LICmQEpylF1emUMpFIJuAPfwj/9t8mMTaN\nj4u517vvRldBSkslPSLJkzMxIam8Eb0OxPbiqacWIPxNDyZTJPjt2yfxtGXJ4tUsZczCbCYxjzwi\nJjFLNQVcboqLVbTILFZtPJIuZooWJOiFuoQZgBETJloBwC6iZKSEs64O7rsv4UtXHbW10frJsbHl\nUeODQRG7P/hAfs/Ohl96Hgp9cOnP4ifVOZVQ/hTk1caNee3tMn709kY33b0bDh9Ow4R63bpoJs/V\nq6lpqzsL0zMt5qStTVy1Y1u/FRXJIDq9T71hA0eenNuYTIu+vmjpR6QFZyxVVZJRsWNH/LGljOAk\n+Folq2K0M/45ezYU7pasCvf6NLz5yhIrWgwMQHkm5c8VFMiCy9GjcmG89x488ED63i8rRqSY1Ngj\nlnlFC8uy3jAMowE4CFjAi5Zl3Z3nZRlBWspDbE6wZ8nNZZlEi9zcXNatW0dPT8+Ur8XYmMyTANzl\nBVRtKIe7d5i8cIGu8Mi1oCyLiT64/h1pcWoYsOkXIXfhuYSx86SurgXOm+7ckREDZGRYwl1rclLK\nU996S2K2CLW1MoalaD63Kqmpkb/N7dsLj8va2qKdNiorU169kzyxKar9/Un5RCTDxo0ytz59Ws7R\nmTMiHiTEsiQH9/XXpX4CRGF49FGx8U8ym+DWLSkHicyLc3NFPIlpgZ0cHk/UpPTatZSYlIIEE5s3\nS430+fMSCM/QHhKZxDidMrA3Ni5zKk4KKS4WYz4VLTKC1RyPpAu3Wy7T/v5ZfC2kF+ri38AWY0gZ\nMsGyyyQb5B733HOrOi09jtpaCRosS5SAHTvS+35jY/CP/xhdQSj0wGf2w+Rr0BXj/+MqhfVPQsH2\nuHM9PCxl9ufPRzctK5M/SdoWEwxDxt8PP5TjXlRKbXJMz7RISHe3ZP1cuxZ9LDdXSmsaGmY/tiwP\nmEOMDg7xfoeEnrFVjBEqK6NCxTQPxtRgWTB2S7IqfOfjy39AsmqK74eCbfHfxTVGbLZzRvlaRHjo\nIVkpHhiQWGfXrvRllzlyo/PMBGac9zKzRtWGYWwP9ylvQIKDyHC4yTCMTZZlnV2WI1wCacm0ADHj\nDE4uW3kIiK9FT08PXV3iV3H8uBN/+O0PHwb7rXp46w4dFy4QysuDrKzkTTjNYWj//6IizIZPQsHW\nRR1nUZF8j/1+SSdfEMeOyQ3cZpOajUUQCMh95ac/jXa1AFHIn3oq/aVoqwGvNxqXdXRI1kky+P3S\nyQJkPvq5z6UtVpmfzZslWgNZ7UmRaAGSgRNZpHzzTSkTKZ2ecDQ4KEtbly9HH6upkTSEGRsnxrJE\nFPnJT6L3p9pa+PmfX4K3Sl2diBadnUt3Wo1h//5ZDDl9PsmOWqsmMZF85OFhUUJXW6bIGmEtxCPp\npLJyDtHi5s2lDXxxmRYmnGqJ5ssfOrQMLaOWkY0bo6WH6RYtBgakPdPdu4AFVVlw0A7DR6PbZHlg\n3SExV4zpAhEKyaLMsWNyTwYZkw8dEtE97eNyRLSYnBR1Pw1m6SCfKeIvPSPTYmBABuhYxSYrS1YZ\nHnpozrHP54PbbQWM3ISeQR8/mzZJXr9ekjN27kzj5R0Yg8FzklUxdif+OWceFO2Tf6419P2ag5wc\n+XubZoaKFg6HeJNFTHLffHOetNMlYBjiazF2WzMtpjHXUuB/BL4M/D9IkDCdlVpjTZq0iRaOHHHx\nXaZMCxDR4p133iEYDPLBB128+64EIV5vuKQ/tx7eeou2/n6YnMRWWRlXVjIrwQm4/rJ8HhAH6uKG\nRR+nYYhAcPnyAs04b96UmjEQg74FTkJDIfHJOX48/oa3bp2IFVu2rJ3FoKWycWO0y0t7e3KihWXJ\n4lokoeATn0h6bp4eysokZW9oSFZYHnwwZbt2OuEzn4G//dtomchv/Ea4TCQUEqXh2LFovZHbLYPZ\nnj1JX2Sjo9KlJGK8b7OJWPLII0u8Tuvq5PgiJqXJKlLzUF8fPd3vvQf3bx0Vk5h33403ibnvPkm/\nWSsmMbERa8blrN5TrPp4JJ1UVoqYODws39GCgkS9UBdJ7Oru3TuyIgDyXXjooaXtO9NwOmWAbG+X\ntMJ00dUF//APMhA4fLB5AraVSCYLSIy57jEo3j+ja1tnp/gexWYF7NghQ9CS2swvhOmliGkSLUA+\n09hYTKbF6Gi0fWlk7LHZRFl/7LFZ62GGh6MeFZ2dUJfjYaMbXLYhwKKszGDXLhEq0hbbWBaMXg+3\nKr0Qb25rGJC/RbIq8utnbVW6VjEMyWTp7c1Q0QJg69Zolm9Li1xz69NUqqOiRUJmFS0sy/py+P9D\ny3Y0KSY2OzvlmRaw7JkWIPe873+/HZfLi2HI5NEwkIE2O5v2wUGYnGTDgQO45ltlDQWh43/IFwNE\nrFj3+JKPNSJa9PeH3cyTcVOO1LrY7ZLWlySRDP1jx+LrOYuLZe60c2dmNyhYCbKypDzmxg0x40yG\n99+PupFv25ayjpqLJ5Ki+v77EmAuqh/a7GzaJDrImTMS4PzsZ/DQppui3MRGi3v3itHmAizD29vh\n+9+PCkCFhWK2WVU19+uSoro6qkhdu5Yy0cJmE2+Lt4+O43r7FEPtZyhwxaSxrlWTmOkdRFS0WBHW\nQjySTmLLHW/dCtfbz+iFukimMi0sOPK63FsMQ4wiVyzVLo14vXKT7uuLKECp3f/HH8sAYPkgvw1q\nc6CuFjCkS1vpQ/JvWsc2v1+y8s7G5BQVF8Ozz0ri4bKSmyvmlt3dMs48+WTa3srjkSF3uG8STpyW\n+tTY9qW7dsn7J0iJGBmR093aKvFObOONiVABbjesWxdk1/N+1lWksU2oOSRtSvvfh8lp7UeyiqKt\nSp3pMMpYPWS8aBGZdH3jG9FOCF/6UnpWRCPmm5Pa9jSWucpDfoHEKxoAWJb1/bQcUQqJHU9T5mkB\n0bany5hpkZOTQ3l5Oa2WYbsLAAAgAElEQVStd+jsvM6+fSLyTcXQNhujGzZwe2QExsfxzqd8WxZ0\n/UBaKIGUg2xITW1q7OTr5s0kOlR0dETrERsakioctCx5yZtvxsdk+fmieezbtzbjqVRRUyOD+J07\n8wtLfX3Rznb5+dIyPiOyViKihWmKspDi2p+nnhLxzdczwfVvvMmeinfIcYdviSUlkhq4gPcMBiUT\nKFKaAyKqfepTKXTKzsoS4aKtLb7Gd6mYJgcm3yH4s7dwmGPcNqFgC/L5n3oqoeJimibNzS20t/sx\nTRtOZwivN4empgaczlVSm6ttTzOCtRCPpJP166PdOm/dCmuVM3qhLnJSFsm0uH0buoYAj5gvpkRl\nzUBqa6OLKG1tIkynAssSs6Sjr4K7HVw9ksJWWSnZFMUHYN2j0fgy5mVnz4pgMTYmjzkcYhW0AOuk\n1FNXJ6LFrVuiqKSpvr8gN0jlzbNsfuc49Me0L/V6ZcEgthQKOZQLFyTz6Pr1eKECJAlw507Y4/VQ\nOgoGgMcHpFi0CAVh+IqUfwxfmdaq1CH+JMX3Q25NhgRUK08k9M9Y0QIky/eBB2RFq6NDVLGdO1P/\nPlNtT0NiFnuPlAnNx1y3u08hQcI64GEgfBfnCeAUkPFBQlo9LWBZMy0ANm6s4dVX7zA+fpOsLJMn\nnogP/NsjZneBALXzGd/dbobBcC1g7kYx3kxROlplZTSA6uqaR7SwrGiA4HSKieE83LghJr4dHdHH\ncnJkAD9wQHajzI3XG83y7eiYvXQ3GIy2NwXxW8iYzna1tdEL7erVlIsWTif8ws4LfPSvr+OaGOKS\nD/Y02LE92ijX6QKixcFBOY+dndF9/9zPibiW8nilrk6C7f5++beUotxgcMokJm94mPICEbGujm2g\n+vNP4dqe2DfHNE1eeukIfX2NZGdHc5a7uny0th7hxRcPrw7hwu2OFlWraLGSrPp4JJ24XJLS3ts7\ni6/FVFuRRWBzgjkpIqgRbpvw1FNLPuaMpbJSTujEhGRcLEK0ME2TllPN+PvbsVkmIezkXL1FQ+g6\nTs9tcNhgx30ifhftkwzXrJm1Hd3dUgoSW2pbXy9jx4pbidTVxZuWpnriZlnw8cdse/MouVfk3hsI\ngKNqvYgVkfEfuT1fvCgZFW1tsggei8cT9aiIxKf4CyDS5MocAuLFj0Uz0QcD70tmhTkS/5y7PNyq\ndDc4Vqk5dRqJiBbj40tvepRWDh0SL5XRUfFW27Il9ROPuLanfSpahJmrPORLAIZhNAM7LMvqDv9e\nAXx7WY5uiRiGpDWHQmnwtADxgwgFZtQcpovBwRrGxn4GBNm2rZOcnPgJQ3v4Tu202agaHU2whzB3\nz0DvKfnZVQrVv5xSV+LsbBEje3qS8LW4di2qPjzwwJzGfbdvi74R632YpO+SMo1k/caOH48GwQ8/\nLHFCxpCTIxHIzZtyHTU1pW7fPh+8/jobLl7EXypv0UE1wQOfZP8TZQvaVWurVJVEutiUl0s5SNnC\ndpM8dXXS+g3kvCwmug2FZKnq2LG4fvfrdq3jRN+T3C3ZSrnfYDb3m+bmFvr7RbAIhST+tNvB5fLQ\n399Ic3MLzz57cOHHtRIUF8sFoKLFirEW4pF0U1kZFS0sC4xUiRaGQ0ThQAAIyYx5LQ+2NpukIl66\nJDPgBbbUNk2TI6+8RGN1H566bDDH4eJb+FydHLlk5/B9dTj37IEND0D5kwnbyifqou3xSGb6tm0Z\nsjC/cWPUOfHq1dSKFu3t0NxMsLOTvsvd3LxpMuTI59rIXio27qRp0yZCk8aUUHHt2swYPz8/KlRU\nVSU4Z7GlGOZsrUmSJGSC74JkVYxcj3/O7gLPrnCr0soM+eNlJrFJ1oOD6bOLWDLZ2VKS9KMfSaz4\n1lupb6WXFRO3qa/FFMnMtjcCsY2A7gCb0nM4qcfhkPK3lJaH2GOWmoN+sKW/Dm14GK5dqwEMcnMt\nCgquA/EzyLbubsjPZ5PDgWM2E6nBj+DWG/KzMw+8X5yRjpgKqqpEtLh5U+Y/CX0lYrMsXC5xIUxA\nX5/MnSLdUEH+rpFuihmz8r+KcDjkb3T9+uy+Fh0dci8GGTzSWLa6eDZvlovs9m0pYF1qU/qILfub\nb07VzXp3unm3oonLuftoPWtQ+2ByOoBpwhtviF9ThAMHpNtPWpMMysvlPIyMSDR34EDyr7UsCdbf\nfDO+UX1RERw6RNnO+5j8UxsMyedqmEW1aG/3Ax7a22W+VFUV9WpzuTzh51cJKlpkEqs6HkknGzZI\nQwe/XwL+oqJEvVAXQUdn9F5QWx12/l7j1NbKfXB4WLp7JKswBydpOf49Giuv4rGHoH8IOj4EcxxP\nrkHjboOW3G0c3PsfMZ1l4fK5c3Hlc+vWNXDsmJPIupPdLosyjz2WYc2LHA4Rd65ckXFmgeJOQm7f\nljqYq1cJBoOcOXONO8PbeCv/GS7l72e7aWfin328/J0jbNlyGMuKH0hzc2UBZtcu8aWa83AceWDY\nwQqGMy0WwdhtMdUcPAfB8fjncjeFW5XukPaVyrzk5ppcudLCwICfiQkbVVUZXFK6b5+YwXZ3i8/K\nvn2p7YnryIu2PVXRYopkRIufAD82DOM7SPnX54HmtB5VComUiKQl0wLE12IZzHOOHgXLcpOXV05d\n3W06Otrjnh8cHGRgYABKSqi12eTmP91EaqQdOsM9K+0uqPkiZKWj8bQEUGfPSobl3buz+PNduhQN\npg4enKE++Hxw4oR0U4yk+9lsYgL52GOp98e61/B6RbDo7Z053x8bC3uFWRKb/MIvrGDt7FzU1clF\nAhI47dmz+H3dugX/8i/xAf7u3TieeYaHe3O5/N9FiPjhD+f3XurpgVdeiZrDut3iBbIs8b5hSNB9\n7pysWAWDyRm8tLfLjSY2PSovT0xi7r8f7HZswO7dJn/7ty28846f7m4bpaXRwMLhcNLZCe++a6Oj\nI7pKeOuWLMxFxEvTXEXuuBGFamgo5YavyoJZ1fFIOpmeWFFURIJeqAvENKH5TTnTdjs0rrFuIbMR\nW2rY3h4VLUIBMcYzB2P+H5CfJwchMIr/0jt4tvphdEIGgkjw6S7Gs/lh/Le2YTrLZpTPjY7Ca6/5\nmJw8wsGDh7HbnXi9YrSZtqy8pVJXJ6LF0NDCxJ3pDAzIytS5c1MPXb3Ry8dlv0z77sNcaJHMntZW\nAA+BQCPQQn39QXJyZFzduVM0lKSN1w0DnPnhv+MCMi2C47L4N3AW/NO+V44cKNorJT/ZmfpHy0xM\n0+Tll4/Q0dGIw+Ght1fWMjO2pNRmk6yzb31L4oIjR+CXfil1+49tezqhokWEZKKv/xn4eSBiNvCX\nlmX9c/oOKbWkRbSYnmmRZrq6ZOIOsGtXDdnZt7l58yaTk5NkhaX3tkhmRXEx3shs6urVaJuHsdvQ\n8V1RlQ07VH8B3OnLvZpuxjlDtAiFolkWbndc67TRWbop7tolGVgrXsu5Rgg3pAFEvNi1S362LJm7\nR1qMPfNMBgdNVVWSqjc+Ltf7YkSLyUm5Fn/2s+gsu7hYjDbD9TA1uZLZ8847koHyzjuJu6xalmQg\nvPFGNLtr0yYRfZatHR1IMHnunKiGN2/KQcxGV5d8/tjsLLdb0pgeeCAuLcQ0TU6digYWFy9KF7CO\nDh9Hjhyhvv4wvb1OenpCU6eyoEBEzFiRx+mcVnScyURuOJYlwXXGfhnuCVZ1PJJOysuj5bC3boWz\n9WN7oQ4Pz1l+mZDjx2FgGIqRiXzOGi4LATFPNH3gHoaiQRgfgGv/ACXnw5Pb4Xl3YSMEY3642wsh\nIJQFxVug/n4wbNgsc6p8zuXyMDEht+CuLrAsmZB3d7fwwgsH2bUrw6sJprc+Xei9MVGwZ7NBQwMn\nCuDW0HM4EgwV2dke8vL8fPGLclku2nTd6Qn/XefJtLAs8N+QrArfx9HWtCB/oLy6cKvSrdrlYZE0\nN7cwPNxIVpaUlEbKaTO6pHTTJmnzfv68GHK2t6fWW03bns5gXtHCsizLMIyzwLBlWc2GYeQYhpFv\nWdb8d+8MILIolpbuIZD2DiKWJRMgkBvzZz9bwxtvnCEUCtHZ2UldeNBob5fMC/e6daw3DFkqv3JF\nRIvJQbj+snhwAGz8echLrWnhdMrKJJVxclIG4337wuZUzc3429uxdXQQ+ugjcoqKaPjyl3FmZzM+\nDqdOiSlvbEerrVulNEG7DaaWDRuiJant7VHR4ty5yIqG+Avt379yxzgvNpsMEhcuLC5F9dIleO21\nqEJjt0uZ0qOPzqjhePpp+UoNDEgGa319vIA2NgavvhptDWsYkhH0+OMr0HZ3ejCZSLTo6RGx4uLF\n6GNZWZL19PDDCV2wmptbGBtrpLzcQ1+f7MLphO5uD2NjjfT2yupXSUkO2dk+vF7PjHnS+PggXu8q\nquma3kFERYsVY7XHI+nE6ZTFgdu3ZzHjvHlzYS2Qb9+WbhfYROyo3CCZBqsZKyQT1NhsiUimhBme\nvEbU1vW35AY3eBNGChKPK4ZNslWdhfJ/ViGhseuEbvgImhsIBh2MVdbhz61i4rrB5CSced/JB21+\n+vs9cXFOhJoaD7t2+bnvvvSeipRQWiqq9NCQjDMHk5xUTk5KoPf22yKsR9i5U4K9khL8V44DMnZW\nVkoiR1GRXONFRVBcbFt6q9dIlvRsooU5AoMfilgx0Rf/XJZHMiqK9qYtY/leor3dT3a2h+xsKXEb\nj6m2yeiS0qYmiaFMU9rs/fZvpy7gi5hvTg7IvStFzRJWM/OKFoZh/Dvgy4jWXgdUAd8AVoV9dNoz\nLdIsWpw7F83Wfvhh2LOnmh//2MCyLK5fv05dXR2WZU1lWtR4vdiysqS49do1mBiG638fXSGoeAYK\nd6X1mEG+sxs2yGS4qytsTvXSSzT29eFxuUSZHBvDNzbG62+fxmM8wpkzzqmWXjBnN0UlBTgckrLf\n1hb1tejvF7dykPrQz3wmw1d6QHwtLlyQka67e0YLtIQMDckAE1EYQCb2n/zkLLVMMp//9KfhW9+S\nusv/8B/8NDTYyMoKkZeXQ29vAyMjInQUFMDnPhefzbKs5OVhlpXRcuoU/suXsX34ISGnkxyvl4b9\n+3G+/bbcXCIBut0u3heNjXP6grS3+3G5PFRWitdMMCjdfAAcDg9+v58nn4QXX2zgr//6CP39jUA0\nxWRiwkdJyds0NR1O44dPLWZ+Pi1XruAfGMA2MEBoyxY5j01NmZWueg+w2uORdFNZGRUtLAuMiooE\nvVCTIBQSk7lQSIzGt24N78ec/7UriWVBYDharjG9fMP0SfCfDEVFcKcHJuwQKMRa52XSKmQ0UMTI\nRCFD44UM+fMZHrExMgLDvhDFLc30fmBSstdOvjuLu2U78PeVQni+Ozw6TtddL4ODthmCRUGBLBLI\n7XeVTE4MQ8bfs2cliJivfC4YlDblx49LTWoEr1dWBTZsmHooNhtvy5aZXeiWmq1nmiYtZy/i73oH\nmxEi9LGdnJI6Gh56Cuf4DSn/GLoUf70YdijYKlkVebU6iUwhkZLRRKJF7PMZR0GBrE4dPSoi53vv\nSYZqKohtezo5mJEdROIWo00zGmemKT5Kpjzkd4AHgDMAlmVdNgwjcVSfgaRHtIhpVZSG8hDTlPTB\ny5f9vPmmjUAgRGVlDg8+2IDb7Wb9+vV0d3dzPTzT7OnpYTTs2lRbWyup3R9+CJNj8P7XISf87S99\nCMqWrya1qkpEi54eOP1aswgWkfZYY2NYFkx66nD9ZIhXPm6mtP5ZQMatp56ScSzjJ8yrHK9XRIu+\nPjFv+/73o1kun/2sCBcZz/SsgrlEi1BIUlHffDO6wpOdLWr5/ffPe8FVVZncvBktj8jOlnvL1as+\nsrOlFnnHDief+czKGsSapsmRDz6gsaMDj8Mh5hqhEL7jxzny0kscfvBBnHa7fN69e6WFVxL1K5HA\nobg4WpUD8tING6CuzsZjjwE4efHFw2GjOX+c0VxTU4bVps6BaZoc+au/orGrC49hyHksKsLX1cWR\n1lYOv/jiqvksa4RVHY+km1gvqb4+KC1N1As1Cd57TzIzAA4+DHmnRRBY6UwLy4LAaEyWRFiUiP3d\nWkCwZxiE7PlMWIWMBQoZNQsZmSxiaKyQkQEb5T/9JpOTNq5e28/l8kdnzdi1ByfZfuH7FNy9SI69\nlqMfD7LrCS9ZOdHOICNjE5xsL2HH3iaGJ04yOSlCeFaWhGxFRdHhZ1WVz9XVyUVnmtLTO1F6vGXJ\nAsHRo3JhRli/XsSKuroZY6/Xm0NXlw+Xa+a4tNRsvakOL+sui/8IQHEbvv7THPn6Nzj82F6czpgy\nD1epCBVFe8CxGoKi1Ufkmi8qkqyx6fFTRn8nHnpIvgMRb5Zdu1ITAE7vIJJhokXcYnRMZm4646Nk\nRIsJy7ImjPANxTAMB9IvfVUQEX1TKlrY7GDPFkOeFGdamKY5ZdB065aHoXDWWijk4xvfEDOampoa\nuru7p3wtIqUhAF6vV2aahgV5H0N3jgwIhfdBxfKubpavm+T2+aO4bl/nzA/+hZrxOwQcIYqK8vD7\nbfSNuLhOJS7DhmugnbIyESsiizpK+tmwIerW/OUv2xgdDVFUlMMXvtBAff0qmYwVFkpgfveuiBaP\nPpp4u9u3ZfUwEoyD1CM+80zSXUeam1soLZXvZsRGA8Bu9zAx0UhxcQtf+MLBFb9+W5qbaXQ4RLAA\nCRgHB/GEQjQGArS0tXHws58Vk5jSmS33ZiMSOBiGnLq+PgkyIiUg2dmhmG2dmVeDukBamptpHBjA\nk58vK4PhVDCPy0Vjfz8tzc0cfPbZFT7Ke4pVHY+km+lmnKWlJOiFOs/NaWhI6t9AvtxPPAGX3xMX\n+1CaMy0sC4Jjs2RJhP9P8hiCIZicgPFgHmPBQkYnCxmeKGRovIhBfyEDo4X0DXkY9TumEs6m8+Bo\nKe6xfpxd7QRKEo8r7uAI+y9/h6LJW7jKgLL1OJ/573T2vYvdbCfHZeLMdlK83sv/9VVZfaypyeHk\nyfRMyJedyOqSZcn4O120uH5dWnDHjruFhVIGct99s16PTU0NtLYemfL+iJCKbL2WU8001vTjcebD\nYPjB/vfxAI21AVo+auNgww7JSi66H3IS9UxVUklEpNq4cRV+JxwOiSO/+12JEY4dg+eeW/p+XSXR\nnyf6YYGWRGkjGITRUVpefZXGa9eiCzqmCeXleHJz0xYfJSNanDAM4w+BHMMwmoD/CfhRSo8ijUQy\nLVLqaQHiaxEcT3mmRcSgaXLSQ2enPFZQABs3Rs1o6upqOH36NKFQiBs3btDW1kYwGGRiuJsPf/pd\n7JiE/CfIGeqjIW8Dzj1PQ9Uy5PlblqRVXL9O4No1bv/dT3jg43XYbNkMm0GCjjzuTgbo7h4gO38d\nd9dtwzJsZGfD/TtNfvnfr0Dt/z2MaZq88soRuroaMQwP/vClHAiIW/MnPrF6VsSpqxPR4sYNWWZ0\nxRjGTU5KOuqZM9E2NEVFMqgssCi2vd1PTo6HrVslmSlCTg7s2OEhK8ufEbGNv70dT1mZeHaEQnHt\nOj3r1uHfuROef37B+41d/crNjc/EyfjAYhH429ulnM3tjhMtQIQLf4xgrCwLqzoeSTfr1kn8HAiI\nRrF7NyJaRHqh+nzzt+V77bWpdDvzmWdoOdWMv/UdbKFxQrm3yakN0vDw3Km/kWzRmVlWDThtwcRZ\nEpH/gxOz7tcCAqYc3uQkTEzCuJnDqFnI8LiUbAz6i+gfkZ/Hg4WEWNwY5nZDsLqWwtv9lLlvkPdg\ngLxCB3l5onHn50OevwfX976DsT08862tlQ4C2dnA7MF6Oifky43pdNLS14f/6lVsra2ELl2S9PD7\n7sP505+KEVSEnBxJpd+/f94uTE5n+rL1/P3teOpcEJhpKuLxFOEfroLtvydd9pRlYdV/J7Zule9/\nW5tkqjU0SCbRUnDkgc0pQu1k3/zbL5awCMHoqIwT8/0czlL2v/MOHv+0OXBeHuTmpi0+Ska0+M/A\nbwHnga8ArwF/k/IjSRNpKQ+BsK9Ff0ozLSYm4MQJPx9/7GE4xlZs82bRGyJmNE8+uRfDMKa8LNra\n2hjqPMMXD+VzYPNdGL0OBwbx3fJx5P0Ahz/9JE5bGtr0hUKyet3RIWr6jRtTQX37lS4YWUdWVjaB\nAPgCLm6F8hikkCFXNq68MdZ7iqivhooK8Jc7VbBYZpqbWxgcbKSkxDM1p7XZYPduDz5fhro1z4JZ\nXU3L3/+9+A587WuEqqokcKquxnnkSNRo02YTo83HHpthtJnU+4TLI4qKxK/ixg0xiN28We41mVJ3\naTNN+axFRdF0XI9HVsEKC7EtMnVx1QcWC8Rmhld13eGSwPHxcJ2/Lf55ZblY1fFIurHb5X5082ZM\nNUiMTwA3b84tWly4MGXMa27fzpGzP6Kxug/PtjERE9yD+GwnOfJKK4efT5z6a06M8ld//gPGfNvw\n5ORSYBskOzDIRGs3pzr+gsaHNmKf1mEhZIEZI0RERInJCRgzXQxPFOHzF+Lzi6fEeLCQ8ZD8C1rJ\nTyxtNhFap0SHvPifYx9zOIBWL7zyHhBg57ZppQ9tbfCP/xitkdu3TzyRkmhlkc4J+XIylR7e2yuT\nF78frl3Dd/QoR0ZGomWITqek0M9i8Dwb6crWs0W8WRx5kFUAgTHIXgfuCnDkYesvVMFimVn13wnD\nkBao3/iGxAhvvAG//utLWyw2DCkJGbuzsLangUBUZFiACLFQbKFpJTuGEZchkI74aM6ZbDj18iPL\nsrYBf5Xyd18G0lIeAtEOIkvMtLAsMao8e1Y6Nnz4oS3OgKa6WjItIpimjezs7Clfi7Nnz3L7xgWe\n2T7GpspqESxGroM7B497hMbaQlr+5Xsc/PKLSzpOQE7irVtRkaKzc9aLvcfvoL+igdbsGi74q+kf\neo+ioZPk2lzYbFCc186DD8r4Pjg+Tk4q2wQpSRExVSwqii7E19ZGKiUy2K15GqZpcuSHP6Txxg08\ndrtcn4CvuZkjo6McPnhQAqeNGyWoXEIbmti6ypoa+X5mYivPUGSA37w5WjBdXDx1sKFFBgCrPrBY\nIFPnKT9fzqHbHSdaLPY8KgtnLcQjy8GGDaJNdHeHL9WEvVATMDEh5sQAbjcthU4a8/vx5GXDZHgi\nbgXx5DporLpJy9G/4+CB3TM6cVy/cJmttlIcZVfEgiIgoUMwCP2963n77W7WlVUxacpbjk84GZqI\nChFjwSIRJMK/Byx34uONwelMTojIyVlgNmds6UNbW1S0eP/9qFEpSF1rY+OCJihrpnyuvx9Peblc\ncABnz0qZRSBAS3s7B7/wBWmhtdB2u2kkZETu24aUf8z6vLKcrPrvRFkZPPAA5ttviznl1avYSksX\nbU5pmiYtH3Tgv/0hNuMsobwOcrLLaajfi3NyMuUixAxsNrlp5uZG/4/5ORQISAZqVpbchB2OuHtg\nOuKjOUULy7IChmFcMgyj2rKsjpS/+zKQtvKQSAeRRWZajI6Kef/Zs1IKFMFmC+FySVbR+vXRBb4I\nkUmR1+ulu7ub8fExPMYtvIUDlFiXYSQ8ImfngLURj9vCf/2DRR0jgYAoKh0d8q+zU2qWEuHxyAyu\nuhpqanj3784x6HuCuzdh8ApQ3MQb46183t1PZZGLnBwDux18ExO8XVLC4aamxR2jsmgiWQEVFVLG\n7HbHL8plStbAfLQ0N9M4NISnpETcRHt64M4dPMGgBE5dXRz8nd+RdL0l1m5MNweL3V0mlUfkeL34\nurrwuN0zSmCWKhKu+sBiAUydx7KyGa1OVWxdXtZCPLIcRHwtTFNii/LyRL1QE3D0KFMmWocP479+\nCk9FeLU50iVh4i70nMQD+K9ehsrOGbvpHzAxzWwGByWONgMORiYLGZkoZGSyEPNaG96tT00JE6aV\nAyS+L7vdUJxAiJguSmRlpaf61XQ6aentxd/Whu3CBUIXLpAzNETD2JgI4dKHntXRnzT1TJXPOZ1y\nLmJWBz0VFfh375aFggwjp9iLb6QLT97MbIrB4XFyivW+riwO85FHOPJnf0ajz4fn7l3pJGK3x5tT\nGsa8GRCmz8eRc6/TuLUdT274vjzsw3cnyJE33BzeGF6MWwjziBAzfs7OnvPGmtPTg+/kSbkHTCNd\n8VEyNQPFQKthGO8Ao+HHLMuyPp3yo0kDaSsPiTgIB/3JmVshonxbmwgVly7FH5PNJiVRVVU54U4E\nc5vRVFWUEuw5zfqsLsYm72BMGPgGbRQVFWFz5kLBdijqhtvd2Pp7JIKZT/WanBSR4vp1ESlu3pxd\n7Ynkx0eEisLCuHPgzBJnq/JyGB6GrCwnFQ+9yJmuZlwD7RS6e/CXlZHj9XJYWweuCBEBzOFIvPiW\nKVkD8zEVOBUXi2gRc816Kivx79olNbQpYLWURzQ0NXGktVVWwWIGFBUJF4aex4xjVccjy8F0M87y\nchL0Qp0Wr3R1SVclkHF9715s7Seiz9vC43OMY6WN8Phg2Ak5PPT6Crl+q5Dvv11Mz+ATIlRMFjJm\n5hErSmRnH6di907y86FiFiEirkRjhUhY+nDiBL7ubo643Rx+8kmcv/qr0ir7HmUq/dtmk9WPri6J\nBWtroaAAW1bWyh7gLDQ83MSRV1rFjDNGuPCNTPB2RwmHn9f7urI4Wk6epHHdOjyRjIfz58Fux2Oa\nNPr9tLS2crCmZv793LpCY10fnti5kRHAk+OkcfsYLdfaOFi1dX7xYQEixEJZifgomSHhfwv/H/tJ\nV41bd/pEi/CKaiggJin22W/OPp9kE77/frS0PkJJiXRa3LNHBmnTbOCll2abFL1F0yM7CVz/Hj0/\n/Qu8rkvYbDYGzRAGTm72mVz3hdh7YD92u51g4RjdrR9x4WovfO1rhCor41OUJiakKD9S7nHrVjTd\ncTqlpfEiRWzNSgJiV6SjreGduOufZXx8kIbHLvLkPbJam6mks6XYcjIVOJWUiCoIcnOur4eSEmwL\nVaPnYLWURzidTpPDE80AACAASURBVA6/+GLC/tkqEiaPnseMY1XHI8tBaalkHkxOypC+bx8iWsT3\nQo2+IBiUUgfLkoDpk5+UVqCxKfK5XjAcYIQ7p9mzMUsqact6gY8u5XPhom3Ko/b8nTfx++/DMERH\n3lgit+NIe8/KyhC/+7vLekoWRVzpw5078mBvLx6Hg0agxevl4D0sWMC09O+6OokNYx7L1PI5p9PJ\n4edfFJPZa+3YLJOQ4SSn2Mvh5/W+riwef3s7nk2bJM1tZEQW0sJ4AH9Pj8ylEhGTCeEfb8NTXQmO\nSSAoXSvzvOAuw+N04vdWwPMvrmhXm5WIj2YVLQzDcAO/DWwGzgHfsixr1bmORZT6tJWHgGRbTBMt\ngkHJpjh7VrpAxbbUcjphxw4RKzZtml4TP3NS5HKOs3vjLQ5uGcPe8R5XrlwhMDlGdnY2fcMhLt0t\n5MJAASXFhQQmArS1tVFbV8u1K5cpuDPIvuFCDl6/DoDvww858v3vc7ihAWdPD7P2+iovnyr1YNOm\npFtCRlgtK9L3MmvlbzQVGOXmwvbtklVUUTGlWKY6cFot5RFOp1PbcaYAPY8rz1qJR5aDyKJ3JFkS\niK/7m+qFGub06eik/LHHpp6LS6F3FoBnJyELBgfgasc4r7+7n4JL8YK33Q67duXQ1+ejosIzI1Ni\nfHyQ2trVIYZPZfBF6rQjsZLHg2fXLvx9aXTzXyVMlc+5XHKOYsbaTC+fczqdHHxc7+tKarGZpnwX\ntm4Vo8JQSL4XYd8HW6Tdb6LMiJhMCNv3J2DzIIQmoH8c7G7ILYUsuefaHEZGtOFd7vhorkyLbwOT\nwEmkd9MO4IXlOKhUkvZMCxBfiyxx5O7tFaEi0mEslspKESp27ZrbQHlqUuS/Bf3vweD5cMsbeb5/\n0MdAcD0t/aW0XBrAsuDYVTvP7JokP8fBwOAA3dfa8Az6ee9aAYeziiRQuXkzapA0NsbB+nrZoWFI\nlBPJoqiunmmmsUBWy4r0vcxa+RvFBU7TTDYzPXBSFCUp1kQ8slxUVopoceeOLNg41q2Leg5M9UJF\nHJhPhMtASkuls1KYSAr9w5v6sQIuenuh9y4MDE3wkwslFGyU1F+7XWxzdu6ELVvAbo9mizocq1cM\nn8rgs9vFE+TOHRlftm4Fm007B6Hlc4oynTjz7oMzF7dCZWUiDs+3n0imm80FpQn2c4+axc4lWmy3\nLOs+AMMwvgm8uzyHlFrS2/JUmBz30xrOquic5kuVnS3xwf33J9myNzghIkV/C4x1xz+XVQjF9/Pe\nSDa9I+O4SiYpK7tCUVERReXl/PhaGy5rgIIcF+uMAPv8tRzekIWzvT0uo8LjdOIHcbuurpauCgto\nQ5Usq2VF+l5mLfyNNHBSlDXPmohHlouIr0UwKL7ElZV2CUBie6FaFvzrv0bNtT/1qanU1GAQOjqc\nTGS/yP/x3WaMiXacholpOZkwvJTWNLF1q3NKqIgPH9aGGB6Xobdtm/g0xIwvmVr6sJxo+ZyixBO3\niDaNhSyiqVlsYuYSLaYKKsKu3Sl9Y8MwioH/AVQD14FfsixrcNo2G4G/A9Yhdat/ZVnWny7kfdJV\nHmLZcxgeli5PP/2xn5uj8c/X1IhQsX37/P6XgAgUfe+B7zwEJ6OPGwbkb4XiBsivExdvx3lgnKys\nLHbGOCiWVkrmRGlZGdtNk4MRU8KhIYlCPB4xSSoowFZSAk8/veTzoCgrjQZOirLmSVs8slyxyHIy\n3YyzsjL8YGwv1I8+ktpVgIYGglXVtF+Fjz+GCxcIe1Q4KSiV1N/pGRVJZYuuYmaUPsRMQjSDL4qW\nzylKlFQtoqlZbGLmEi12G4YxHPO7O+Z3y7KsuZ0Y5+cPgGbLsv7IMIz/HP79D6ZtYwJftSzrA8Mw\n8oAWwzCaLcu6kOybpDrTwu+XVqXn3s9hSziMMoJSB5KXJ6ZX+/aJAdW8BCfB95GUgPintSLL8kj/\n6OJ9Uk8ag9frpaurC1cCJW98fByv10uovV0ecDgStuPSVQJlLaGBk6KsadIZjyxLLLKcFBeLqDA+\nLjrF/v3E90K9cQN+/GNCIeifzOPM+NO0/jFTZpoRFiJUrDU0g09RlIWSqkU0NYtNzKyihWVZqbPc\nT8yngcfDP38bOM60QMGyrNvA7fDPI4ZhXAAqgQWLFpYliws228IP1LKgvV3KPy5ciAgg2dQXG9hs\nFrWb/DzaIM0Kktr/2G0p/xg8J+UgEQwD8uuheD/kb472Rp9GU1MTra2t9Pf3xwkXExMTlJSU0NTU\nREtzc0pSlBRFURRlJUlzPLIsschyEgiY9Pa2cOGCn48+stHVFaIuf4T8S5cYG/QxfvIsw74A/kAR\n/gd/j77WqIfVvSxUxKIZfIqiLIZULaKpWexMVrALNuWWZYUtq7kDlM+1sWEYNcA+4GcLeZNY9+pg\ncGGixdAQfPCBtCodGIh/rrjYoLouhw3lo7gq/LAh8T6mCJkw+JGIFf6u+Oec+VB8v2RWZM1sPzkd\nMVB8kebmZtrb2zFNE6fTidfrpSk8mOoqgaIoiqLMy7LEIsuFaZq89NIRbt5sxO/3MDYGXV0mLcf+\nK19pP099TjGWJYHQnSwH37nWTFHFrjk8Ku5dNINPURQlc0iraGEYRjOQyH7yD2N/sSzLMgxj1l7r\n4XTMfwJesCxrJNE2X/va16Z+PnToEIcOHQKimRYg9g7ziePBIFy+LELFlSvxHUEdjmir0upqMK7k\nwPgoBEZn3+F4j3hVDJ6D4Hjsh4K8zeJVUbBl1qyK2ZCa0dkHU10lUBRFufc4fvw4x48fX+nDyCiW\nMxaB2eOR5aC5uYX+/kZKSjzcvi0xzJnXmvnFgXEmzU2Mjt4mJ6cAy2Yjf9sOfqdsgO69zTz+WZ2c\nK4qiKKkj1fGIYVmzjs9pxTCMi8Ahy7JuG4ZRARyzLGtbgu2cwL8Ar1uW9dIs+7Jm+xzvvisG2QC/\n93viO5GIvr5oq9KRaaHI+vUiVNx3X7QTqGmatHzvf8F/9xI2l4dQaSM5xV4aHm7CaQd8H4tXxei0\ndiLOvLBXxf1TbVIVRVEUJR0YhoFlWSvf0D1DSWUsEt5u1nhkOfj619+kt/dJxsfhzBl5rODm1/mc\n2cuGsatsNM5TVVVE9s46HN6NALxZVsaTv/M7K3bMiqIoytpnqfHISpaHvAr8OvB/h///wfQNDLEI\n/ybw8VxBwlzEZlpMN+M0TXHKPntWeprH4nJJq9J9+6CiQhIjoq8zOfLKSzQWteEp8ctZLOnFN3iZ\nI9/8HocfqsNpTMbvML8u7FWxBWzptgtRFEVRFCUJliUWWS5MU7I2XS5Yt04WZArdJqUeyHZU4O6/\nQF51CdRUTb3GFml7qiiKoigZykqKFv8N+EfDMH6TcJsxAMMwKoG/tizrOeAR4IvAOcMw3g+/7r9Y\nlvVGMm9gmianT7fwzjt+QiEbf/mXIXbtymHXrgbOn3dy7hxMTMS/prpasip27Ji9lKTlVLO0oQnl\nwJgPQhMw8D6eSR+NFQFaPhjn4L56cORK94/iBsgqWvAJUhRFURQlraQ9FllOnM4QIAstO3bIY8Pv\nOMnzQ5Bc7mzeSfUub9xKjHYTUxRFUTKdFRMtLMvqB55O8Pgt4Lnwz28Bi+j3ETWjunhRzKgAzp+H\no0d9hEJHOHjwMHa7DNS5ubB3r2RVlJbOv29/fzueOheMhAf6UBAmfQB4chz4e+yw6Xko2KZZFYqi\nKIqSoaQ7FlluvN4curp8uFxRU++JIi/DQ124sSgqyooTLLSbmKIoirIaWMlMi7QSMaNyu6MDd2cn\ngIdAoJG2thaee+4g+/aJW7Z9AdqCzQqnUtrdMQ86wb0e3JXYhiugcGdKPoeiKIqiKEoyNDU10Np6\nhP7+xinhoqi2iTduvcenOU9tbdSuQ7uJKYqiKKuFNStatLf7cbmk3Vcs2dlQUeFh1y4/v/Iri9t3\nyAhnWGSXAxYYDnCVTnUAmXpeURRFURRlmZCW6Idpbm6hvd2PadpwOkMc+OqT5LKXE52d2k1MURRF\nWXWsWdEiYkZVUACVlRAKQXk5FBZKZqTdvvhMz5xiL76RLjx5LnBXxj03ODxOTrGmWiqKoiiKsvxI\nS/SDK30YiqIoipIyVkWN5mKImFHZbFL+sW0bFBVFSzkjzy+GhoebeOt6Mb6ReBdP38gEb3eU0PCw\nploqiqIoiqIoiqIoylJZs5kWicyoIoyPD+L15ix6306nk8PPv0jLqWb819qxWSYhw0lOsZfDz2uq\npaIoiqIoiqIoiqKkAsOyrJU+hiVjGIY1/XNEuofEmlEBTEz4KC5+ixdfPKzigqIoirKmMQwDy7KM\n+bdUUkGieERRFEVR7nWWGo+sWdECRLiYbkbl9ebQ1NSggoWiKIqy5lHRYnlR0UJRFEVRZqKiBRok\nKIqiKEoiVLRYXjQeURRFUZSZLDUeWbNGnIqiKIqiKIqiKIqirG5UtFAURVEURVEURVEUJSNR0UJR\nFEVRFEVRFEVRlIxERQtFURRFURRFURRFUTISFS0URVEURVEURVEURclIVLRQFEVRFEVRFEVRFCUj\nUdFCURRFURRFURRFUZSMREULRVEURVEURVEURVEyEhUtFEVRFEVRFEVRFEXJSFS0UBRFURRFURRF\nURQlI1HRQlEURVEURVEURVGUjERFC0VRFEVRFEVRFEVRMhIVLRRFURRFURRFURRFyUhUtFAURVEU\nRVEURVEUJSNR0UJRFEVRFEVRFEVRlIxERQtFURRFURRFURRFUTISFS0URVEURVEURVEURclIVLRQ\nFEVRFEVRFEVRFCUjUdFCURRFURRFURRFUZSMREULRVEURVEURVEURVEykhURLQzDKDYMo9kwjMuG\nYRwxDKNwjm3thmG8bxjGj5bzGBVFURRFWdtoPKIoiqIomc9KZVr8AdBsWdYW4Gj499l4AfgYsJbj\nwBRFURRFuWfQeERRFEVRMpyVEi0+DXw7/PO3gc8m2sgwjCrgWeBvAGN5Dk1RFEVRlHsEjUcURVEU\nJcNZKdGi3LKsO+Gf7wDls2z3J8DvA6FlOSpFURRFUe4lNB5RFEVRlAzHka4dG4bRDKxP8NQfxv5i\nWZZlGMaMVEvDMD4J9FiW9b5hGIfme7+vfe1rUz8fOnSIQ4fmfYmiKIqirCmOHz/O8ePHV/owMgqN\nRxRFURRleUl1PGJY1vKXZhqGcRE4ZFnWbcMwKoBjlmVtm7bNfwV+DQgA2UAB8D3Lsv5Ngv1ZK/E5\nFEVRFCWTMQwDy7K0nGEWNB5RFEVRlPSz1HhkpcpDXgV+PfzzrwM/mL6BZVn/q2VZGy3L8gJfAN5M\nFCAoiqIoiqIsEo1HFEVRFCXDWSnR4r8BTYZhXAaeDP+OYRiVhmH86yyv0aULRVEURVFSicYjiqIo\nipLhrEh5SKrRdExFURRFmYmWhywvGo8oiqIoykxWa3mIoiiKoiiKoiiKoijKnKhooSiKoiiKoiiK\noihKRqKihaIoiqIoiqIoiqIoGYmKFoqiKIqiKIqiKIqiZCQqWiiKoiiKoiiKoiiKkpGoaKEoiqIo\niqIoiqIoSkaiooWiKIqiKIqiKIqiKBmJihaKoiiKoiiKoiiKomQkKlooiqIoiqIoiqIoipKRqGih\nKIqiKIqiKIqiKEpGoqKFoiiKoiiKoiiKoigZiYoWiqIoiqIoiqIoiqJkJCpaKIqiKIqiKIqiKIqS\nkahooSiKoiiKoiiKoihKRqKihaIoiqIoiqIoiqIoGYmKFoqiKIqiKIqiKIqiZCQqWiiKoiiKoiiK\noiiKkpGoaKEoiqIoiqIoiqIoSkaiooWiKIqiKIqiKIqiKBmJihaKoiiKoiiKoiiKomQkKlooiqIo\niqIoiqIoipKRqGihKIqiKIqiKIqiKEpGoqKFoiiKoiiKoiiKoigZiYoWiqIoiqIoiqIoiqL8/+3d\nfXBU9b3H8fc3JATChpBgePKJKAWLc1UMUnUIRWoYxGqBUaud66jVQq3yoK0t2hEd7rU6FW9D69ip\nV9tpO63W1g5atUIQoVEcI1HRBiziDTYBRSAkEkOSjfndP3YJmzXPu3vOZvN5zTDunt39ne9+Oez5\n+Nuz5yQlTVqIiIiIiIiISFLSpIWIiIiIiIiIJCVNWoiIiIiIiIhIUtKkhYiIiIiIiIgkJV8mLcws\nz8xKzWyXmW0ws1FdPG+Umf3FzHaa2Q4zO9/rWkVERCQ1KY+IiIgkP7+OtFgJlDrnJgMvhe93Zi3w\ngnPuy8BZwE6P6hMREZHUpzwiIiKS5Mw55/1Kzd4Dvuqc229m44DNzrkzop6TA7zlnDutF+M5P96H\niIhIMjMznHPmdx3JSnlEREQk8WLNI+nxLKYPxjrn9odv7wfGdvKcAuCAmf0GOBuoAJY75xp7u5Lm\n5mYuvfRStmzZQmtra8xFi4hIfKSlpXHqqady3333cc011/hdjgxeyiMiIoOY8sjAkLBJCzMrBcZ1\n8tCPI+8455yZdfa1RDpwLnCrc+4NMyshdNjmqs7Wd++997bfnj17NrNnz2b16tUEAgE+/fRThg8f\n3s93IiIi8dbS0sK2bdtYtGgRTU1N3HDDDX6XlBI2b97M5s2b/S4jqSiPiIhIVyLzSEtLC9ddd53f\nJaWEeOcRP38eMts597GZjQde7uRwzHHAa865gvD9mcBK59zXOxmv08Mxx44dy2uvvcZpp/V4RKeI\niPhg69atXHHFFVRUVDB+/Hi/y0k5+nlI95RHREQEjueRyspKcnNz/S4n5cSaR/w6EeezwLFprOuA\nddFPcM59DFSb2eTwoouByr6s5MCBA5xyyimx1CkiIgk0ffp09u/fz7p162hubva7HBl8lEdERKQ9\njzzzzDPo3ETJx69JiweAYjPbBcwJ38fMJpjZ8xHPWwr8wcy2Ezpb90/6shLnHOnpfp22Q0REejJ0\n6FDa2toIBoM0NDT4XY4MPsojIiLSnkfq6upoaWnxuxyJ4sse1DlXS+ibiujl+4BLI+5vB87zsDQR\nEfFJW1ub3yXIIKM8IiIikcI/Y/C7DIni15EWEkfXX389d999t99lDDrqu3/Ue3+o7yLSHX1G+EN9\n94f67h/1fvAZdMcqBoNBSksrqKpqJBhMIyOjjYKCLIqLC8nIyPBkjEAggFnoPCSfffYZw4YNY8iQ\nIQA8+uijfb7cjpm1j9eTp556ipKSErZv386MGTN4+eWX+7Su/goGg1SUltJYVUVaMEhbRgZZBQUU\nFhf3qe+xjOFn33/wgx/w7LPP8vHHH3PiiSdy1113ce211/Zpff0V2l5LqaqqIhgMkpGRQUFBAcW9\n7Fusrwd/e//DH/6QJ598kvr6enJzc1myZAl33nlnn9bXH8FgkIqtpTTWVpHmgrRZBll5BRRe2Mdt\nPoYx/Oz7MbW1tUyZMoUzzjiDsrKyPr1WJJUpjyiPwODJI/HIErGOMRizCCiPHKM8MrANqkmLYDBI\nSckGDh2aybBhOe3La2rqqazcwIoVc3v8hxePMSJ/t11QUMDjjz/OnDlz+vmuQnp7GNPo0aO5/fbb\n2blzJ5s2bYppnb0VDAbZUFLCzEOHyBk2rH15fU0NGyormbtiRa/6HusYfvY9EAjw3HPPMXnyZMrL\ny5k3bx6TJk3iggsuiGn9PQltryUcOnSIYRF9q6mpobKykhU99C3W1x/jZ+9vvPFGVq1aRSAQYN++\nfcydO5czzjiDhQsXxrT+7gSDQTb8uYSZpx4i5/SI7bWhhg1/rmTulb3c5mMcw8++H/OjH/2IqVOn\n6lBLkQjKI8ojxwyGPBKPLBGPMQZbFgHlkUjKIwPboPp5SGlpBbW1HXfuAJmZOdTWzqS0tMKTMbpS\nXl7OBRdcQG5uLhMmTGDp0qUEg8H2x2+77TbGjh1LTk4OZ511Fjt27PjCGEeOHOGiiy5ixYoVna7j\na1/7GldccYWnlxasKC1lZm1th507QE5mJjNra6koLfVkjK540fd7772XyZNDJ56fMWMGRUVFvPba\na/2uubdKS0upra3tsIMHyMzMpLa2ltIe+hbr63viRe+nTJlCIBAAQju4tLQ0Pvjgg5jq7knF1lJm\nTqwlJztqew1kMnNiLRVbe7HNx2GMrnjRdwhdPqyyspIbbrhBIUEkgvKI8khnUjWPxCNLJDKPpGoW\nAeWRY5RHBr5BdaRFVVUjmZmhnfvu3dDxRPU5VFc38skn3Y+xcWMjDQ0dA0IgAJMmhYJCVVVjv+tL\nT09n7dq1TJ8+nerqai655BIeeeQRli9fzvr16ykrK+P9999n5MiR/Otf/yIn53gdZsahQ4e45JJL\nmDdvHqtXr+53HfHWWFVFTmZm6E5U43OAxupqemp848aN5ERfWSDc+JzMTBqrqvpdn9d9P3r0KG+8\n8Qa33HJLv2vuraqqKjLDvd+9e/cXrs5QXV3NJ930fuPGjV94TSAQYNKkSWRmZlIVQ9/Bu94/8MAD\n3HfffXz22WecdtppfOtb34qp7p401laRc3p4mz+yG1qjtvld1XByD9v8ro3kTG6AwxEL0wOQPYmc\nQCaNHyT3Nv/555+zdOlSHnvsMbZv397vWkVSkfKIP5RHOvIqj8SaRSCxeSRVswgoj4DySKoYVJMW\nweDxA0saGqCuruPjTU1p7NnT/RgHD6bR1NS7dfTVueee23771FNPZfHixWzZsoXly5eTkZHBkSNH\n2LlzJ+eddx5Tpkzp8Nq9e/cye/Zsrr/+er7//e/3u4ZESIuYMe2s8WlNTfTU+LSDB+mu8R3W0Ude\n9/273/0u55xzDnPnzu13zb0VOVvd0NBAXVTvm5qa2NNN7w8ePEhTN30PxtB38K73K1euZOXKlbz9\n9tusW7eOkSNHxlR3T9JcRF9aG6AlaptvboKGPd2P0XwQWrrZ5l1yb/M///nPOf/885k2bZpCgkgU\n5RF/KI905FUeiTWLQGLzSKpmEVAeAeWRVDGoJi0yMo5fTi98hFYHgUAbEyd2P8bu3W10NsHe2Tr6\nateuXdx+++1UVFTQ2NhIa2sr06dPB2DOnDnceuut3HLLLXz44YcsWrSINWvWkJ2djXOO559/nuzs\nbJYsWdLv9SdKW+Tv3DppfFsgQE+Nb/viV1Edxmrr5YmEOuNl3++44w527Njh2QnHIn9jGOik94FA\ngInd9L6zb0Qix+ntCZy64vU2f84557B+/XruueceHnrooZhq706bRfQlvZNtPjMAgYndj5G5G4ZG\nbfMRY3VYRx8luu/79u3jF7/4BRUV/T88XSSVKY/4Q3nkOC/zSKxZBBKbR1I1i4DyiPJI6hhUkxYF\nBVnU1NSTmZnDpEkdH2tqqmPWrCzmz+9+jDFjsigrq28/rDN6jIKCrH7Xd/PNN1NYWMif/vQnRowY\nQUlJCU8//XT740uXLmXp0qUcOHCAq666igcffJDVq1djZnznO9/h8OHDzJ8/nxdffJGsrO7r6OsZ\nd2ORVVBAfU1N6JDMqMbXNTWRNWsWPTU+a8wY6svKjh/WGT1GQUG/6/Oq7/fccw/r169ny5Ytne60\nE6GgoICamhoyMzOZFNX7pqYmZs2axfxuej9mzBjKysraD+uMfn1BDH0Hb7f5Y4LBYMJ/R5qVV0B9\nQw05gUzIjtrmjzSRNXkWnNbDNl89hvqMstAYUeqONJGVl7zbfHl5OR999BFTp04FQocgHz16lAkT\nJrB3715PP39EkpHyyHHKI8elah6JNYtAYvNIqmYRUB5RHkkdg+pEnMXFheTlvUJzc32H5c3N9Ywe\n/SrFxYWejNGVhoYGsrOzycrK4r333uOXv/xl+z+mbdu28frrrxMMBsnKyupwqaBjJ5R5+OGHmTJl\nCpdddlmXh9C1tbXR1NREMBikra2N5ubmmA/x70lhcTGv5OVR39zcYXl9czOvjh5NYXGxJ2N0xYu+\n33///TzxxBOUlpaSm5vb71r7qri4mLy8PJqj+tbc3Mzo0aMp7qFvsb6+J4nuvXOOX/3qV9TV1eGc\no7y8nEceeYSLL744prp7UnhhMa/syaO+IWp7bWjm1Q9HU3hhL7b5OIzRlUT3ff78+Xz44Yds376d\n7du3s3r1aqZNm8bbb7+tgCCC8ggoj3QmVfNIPLJEIvNIqmYRUB5RHkkdg2rSIiMjgxUr5lJUtJP8\n/E2MGrWZ/PxNFBXt7NWlweI1RlfWrFnDH//4R0aOHMnixYu5+uqr2x/79NNPWbx4MXl5eUycOJET\nTjiBO+64A+h4reJHH32Uk046iQULFnzhgx3gd7/7HVlZWXzve9+jrKyM4cOHJ/wQzoyMDOauWMHO\noiI25eezedQoNuXns7OoqFeXBovXGF3xou8//vGPqa6uZtKkSWRnZ5Odnc0DDzzQ75p7K7S9rqCo\nqIj8/HxGjRpFfn4+RUVFvbo8WKyv74kXvV+3bh2nn346I0eO5Nprr2XZsmXceuutMdXdk4yMDOZe\nuYKdwSI2fZDP5t2j2PRBPjuDRb26NFi8xuhKovs+dOhQxowZ0/4nJyenfZmIKI+A8khnUjWPxCNL\nJDKPpGoWAeUR5ZHUYalw2Rczc529DzPTZW1ERJKcmbF27VquueYa8vPz/S4npYT3g/o6ySPKIyIi\nA5eZUVJSwpIlS75weV2JTax5ZFAdaSEiIiIiIiIiA4cmLUREREREREQkKWnSQkRERERERESSUkpP\nWpgZra2tfpchIiJdaGlpIS0tpXdFIsojIiJJTnkkuaX030x+fj7//ve//S5DRES6sG3bNsaNG+d3\nGSIJpTwiIpLclEeSW0pPWtx0000sX76co0eP+l2KiIhEaGlpYevWrSxYsICFCxcC6BsOSVk33XQT\ny5YtUx4REUkykXlkwYIFOOfaL6cqySPd7wISadWqVcyZM4fs7Gw+//xzv8sREZGwtLQ0xo0bx1VX\nXcXZZ59Na2srgUDA77JEEmLVqlXMmjVLeUREJMlE5pEzzzyT4cOHM3ToUL/LkiiWCtcN7+q66ABt\nbW1s3ryZ5Ex9+wAACM5JREFUd955R9/iiYgkofT0dBYuXMj48eP9LiXlxHpddOmb7vJIMBjkhRde\nYM+ePfoWT0QkCY0YMYJFixaRm5vrdykpJ9Y8kvKTFhCauDh8+LAOy4yD8vJyZsyY4XcZg4J67S31\n2zuRvR4yZAg5OTlkZWX5XFVq0qSFt3rKI62trdTV1dHU1ORhValJn9neUa+9pX57J7LX6enpjBo1\nimHDhvlcVWqKNY+k9M9DjklLS2P06NF+l5ESHnvsMRYtWuR3GYOCeu0t9ds76rUMVunp6Zxwwgl+\nl5ES9DniHfXaW+q3d9TrgUO/lxARERERERGRpKRJCxERERERERFJSilzTgu/axAREUlGOqeFd5RH\nREREOjfoT8QpIiIiIiIiIqlHPw8RERERERERkaSkSQsRERERERERSUqatBARERERERGRpDSgJy3M\nbJ6ZvWdm75vZj/yuJ9WY2clm9rKZVZrZP81sWXh5npmVmtkuM9tgZqP8rjVVmNkQM3vLzP4Wvq9e\nJ4CZjTKzv5jZTjPbYWZfUa8Tw8xuC39+vGtmfzSzTPU6fszs12a238zejVjWZX/N7M7wPvM9M5vr\nT9WpR3kkcZRF/KE84g3lEe8ojyRWovPIgJ20MLMhwMPAPGAqcI2ZfdnfqlJOELjNOXcmcD5wS7jH\nK4FS59xk4KXwfYmP5cAO4NgZctXrxFgLvOCc+zJwFvAe6nXcmdmJwFKg0Dn3H8AQ4GrU63j6DaH9\nYKRO+2tmU4FvEtpnzgMeMbMBmwOShfJIwimL+EN5xBvKIx5QHvFEQvPIQA4rM4Ddzrk9zrkg8CTw\nDZ9rSinOuY+dc2+HbzcAO4ETgcuB34af9ltggT8VphYzOwmYDzwGHLskkHodZ2aWAxQ5534N4Jxr\ndc7Vo14nSjqQZWbpQBawD/U6bpxzZcDhqMVd9fcbwBPOuaBzbg+wm9C+VGKjPJJAyiLeUx7xhvKI\n55RHEijReWQgT1qcCFRH3K8JL5MEMLOJwDTgdWCsc25/+KH9wFifyko1PwPuANoilqnX8VcAHDCz\n35jZm2b2v2Y2AvU67pxze4GHgH8TCgd1zrlS1OtE66q/EwjtK4/RfjM+lEc8oiziGeURbyiPeER5\nxDdxyyMDedLC9fwUiQczCwBPA8udc0ciH3POOfR3ETMz+zrwiXPuLY5/q9GBeh036cC5wCPOuXOB\nz4g6HFC9jg8zyyU0yz6R0A4qYGb/Gfkc9TqxetFf9T526qEHlEW8oTziKeURjyiP+C/WPDKQJy32\nAidH3D+ZjjM2EgdmlkEoJPzeObcuvHi/mY0LPz4e+MSv+lLIhcDlZlYFPAHMMbPfo14nQg1Q45x7\nI3z/L4RCw8fqddxdDFQ55w4551qBvwIXoF4nWlefG9H7zZPCyyQ2yiMJpiziKeUR7yiPeEd5xB9x\nyyMDedJiG/AlM5toZkMJnczjWZ9rSilmZsDjwA7nXEnEQ88C14VvXwesi36t9I1z7i7n3MnOuQJC\nJwba5Jy7FvU67pxzHwPVZjY5vOhioBL4G+p1vH0InG9mw8OfJxcTOrGbep1YXX1uPAtcbWZDzawA\n+BJQ7kN9qUZ5JIGURbylPOId5RFPKY/4I255xEJHagxMZnYJUELoDLCPO+fu97mklGJmM4F/AO9w\n/JCdOwltVE8BpwB7gKucc3V+1JiKzOyrwPedc5ebWR7qddyZ2dmETjA2FPgAuIHQ54h6HWdmdi+h\n/4lrBd4EbgKyUa/jwsyeAL4KnEDo96KrgGfoor9mdhfwbUJ/H8udc+t9KDvlKI8kjrKIf5RHEk95\nxDvKI4mV6DwyoCctRERERERERCR1DeSfh4iIiIiIiIhICtOkhYiIiIiIiIgkJU1aiIiIiIiIiEhS\n0qSFiIiIiIiIiCQlTVqIiIiIiIiISFLSpIWIiIiIiIiIJCVNWogMAmY22szeCv/5yMxqwrePmNnD\nHtVwtpld4sW6REREJPkoj4hIf6T7XYCIJJ5z7hAwDcDM7gGOOOf+x+MypgGFwN89Xq+IiIgkAeUR\nEekPHWkhMjgZgJnNNrO/hW/fa2a/NbN/mNkeM1toZj81s3fM7O9mlh5+XqGZbTazbWb2opmN+8Lg\nZlea2btm9nb4uRnAauCb4W9UrjSzEWb2azN73czeNLPLw6+93syeMbOXzWyXma0KLx9hZs+Hx3zX\nzK7yqlkiIiKSEMojItIjHWkhIpEKgIuAM4HXgEXOuR+a2V+BS83sBeAXwGXOuUNm9k3gPuDGqHHu\nBuY65z4ys5HOuaCZ3Q0UOueWAZjZT4CXnHPfNrNRwOtmtjH8+vPCNRwF3jCz54GJwF7n3KXh149M\nWBdERETET8ojItJOkxYicowD/u6c+9zM/gkMcc6tDz/2LqGd9GRCO++NZgYwBNjXyVivAr81s6eA\nv4aXWfjPMXOBy8zsB+H7mcAp4To2OOcOA4QDykzgBWCNmT0APOeceyX2tywiIiJJRnlERDrQpIWI\nRGoBcM61mVkwYnkboc8LAyqdcxd2N4hz7mYzmwFcClSYWWEXT13knHs/coGZfSXqOQa0OefeN7Np\n4TH/28xecs79V6/fmYiIiAwUyiMi0k7ntBCRY6znp/AvIN/Mzgcwswwzm/qFgcxOd86VO+fuAQ4A\nJwGfAtkRT1sPLIt4zbSIOorNLNfMhgPfAF41s/FAk3PuD8Aa4Nw+v0MRERFJdsojItKBjrQQGZxc\nxH87u03UbQAX/i3oFcDPzSyH0GfIz4AdUc/9qZl9idAOf6Nz7h0zqwZWmtlbwE+A/wJKzOwdQhOo\n/wdcHl5vOfA0oXDxe+fcm2Y2F3jQzNoIfQNzc2wtEBEREZ8pj4hIj8y56M8BERH/mNn1hE6QtdTv\nWkRERGRwUh4RSR76eYiIJJvob1hEREREvKY8IpIkdKSFiIiIiIiIiCQlHWkhIiIiIiIiIklJkxYi\nIiIiIiIikpQ0aSEiIiIiIiIiSUmTFiIiIiIiIiKSlDRpISIiIiIiIiJJ6f8Be2lFWfwvVJcAAAAA\nSUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x199a1470>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, axarr = plt.subplots(2, 2, figsize=(18,12))\n",
    "coord = [ (0,0),(0,1),(1,0),(1,1) ]\n",
    "def run(repetitions, graph=False):\n",
    "    # Number of arms or tasks\n",
    "    numBandits = TASKS\n",
    "\n",
    "    # Store collective history of successes and failures\n",
    "    combSF = []\n",
    "\n",
    "    # For how many steps to run the exercise?\n",
    "    steps = 100\n",
    "\n",
    "    # Store counts and total rewards\n",
    "    cRnd, trRnd = [], []\n",
    "    cGre, trGre = [], []\n",
    "    cEG, trEG = [], []\n",
    "    cTS, trTS = [], []\n",
    "    cUCB, trUCB = [], []\n",
    "\n",
    "    # Store exploration proportions\n",
    "    ExpRnd, ExpGre, ExpEG, ExpTS, ExpUCB = [], [], [], [], []\n",
    "    # Store probabilities of success\n",
    "    ProbRnd, ProbGre, ProbEG, ProbTS, ProbUCB = [], [], [], [], []\n",
    "    # Store evolution of rewards\n",
    "    rewRnd, rewGre, rewEG, rewTS, rewUCB = [], [], [], [], []\n",
    "\n",
    "\n",
    "    # strategies = ['random','greedy','eGreedy','thompson','ucb']\n",
    "    # names = ['Rnd', 'Gre', 'EG', 'TS', 'UCB']\n",
    "    # plotNames = ['Random', 'Greedy', 'e-Greedy', 'Thompson Sampling', 'UCB1']\n",
    "    strategies = ['random','greedy','eGreedy', 'ucb']\n",
    "    names = ['Rnd', 'Gre', 'EG', 'UCB']\n",
    "    plotNames = ['Random', 'Greedy', 'e-Greedy', 'UCB1']\n",
    "    \n",
    "    #f, axarr = plt.subplots(2, 2, figsize=(18,12))\n",
    "    #coord = [ (0,0),(0,1),(1,0),(1,1) ]\n",
    "    selRnd, selGre, selEG, selUCB = [], [], [], []\n",
    "    for st in range(len(strategies)):\n",
    "        print(plotNames[st])\n",
    "        # How many times to repeat it\n",
    "        for rep in range(repetitions):\n",
    "\n",
    "            #print(plotNames, \" ==============\")\n",
    "            stats, actions, probs, sf, counts, expSpace, cpTask = simulate(numBandits, steps, strategy=strategies[st])\n",
    "\n",
    "            combSF.append(sf)\n",
    "            #print(\"Total reward and action selected:\\n\",stats)\n",
    "            #print(\"Total reward and action selected \", stats[-1])\n",
    "            # explorationDyn(actions, plotNames[st])\n",
    "            #probDyn(probs, plotNames[st])\n",
    "            #succFailDyn(sf, plotNames[st])\n",
    "            var = ''.join(['c',names[st]])\n",
    "            vars()[var].append(counts)\n",
    "            var = ''.join(['tr',names[st]])\n",
    "            vars()[var].append(stats[-1][0])\n",
    "            var = ''.join(['Exp',names[st]])\n",
    "            vars()[var].append(expSpace)\n",
    "            var = ''.join(['Prob',names[st]])\n",
    "            vars()[var].append(probs)\n",
    "\n",
    "            #lpDyn(probs, plotNames[st])\n",
    "\n",
    "            var = ''.join(['rew',names[st]])\n",
    "            vars()[var].append(np.array(stats)[:,0])\n",
    "\n",
    "            var = ''.join(['sel',names[st]])\n",
    "            vars()[var].append([np.shape(cpTask[i])[0] for i in range(numBandits)])\n",
    "\n",
    "        var = ''.join(['sel',names[st]])\n",
    "        avgSelection = np.mean(np.array(vars()[var]), axis=0)\n",
    "        tasksT = [\"Task 1\",\"Task 2\",\"Task 3\",\"Task 4\"]\n",
    "        for ta in range(len(tasksT)):\n",
    "            print(''.join([tasksT[ta], ' ', str(avgSelection[ta])]))\n",
    "        print(\"Total reward \", stats[-1][0])\n",
    "\n",
    "        plotSelection(cpTask, coord[st], plotNames[st])\n",
    "        print(\"===============================================\")\n",
    "        \n",
    "run(1, True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": true
   },
   "source": [
    "In the plots above we observe an example of a trial for each of the strategies. Since only one task is selected at each time step, the circles represent when that specific task was selected.\n",
    "\n",
    "The difference in cumulative reward is very small, and possibly a product of the general ignorance regarding the stimuli during the very early stages."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 172,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Random\n",
      "Task 1 26.18\n",
      "Task 2 25.74\n",
      "Task 3 25.31\n",
      "Task 4 26.77\n",
      "('Total reward ', 2.8371148459383746)\n",
      "===============================================\n",
      "Greedy\n",
      "Task 1 17.69\n",
      "Task 2 24.6\n",
      "Task 3 32.03\n",
      "Task 4 29.68\n",
      "('Total reward ', 3.040801564027368)\n",
      "===============================================\n",
      "e-Greedy\n",
      "Task 1 17.86\n",
      "Task 2 24.94\n",
      "Task 3 31.85\n",
      "Task 4 29.35\n",
      "('Total reward ', 2.854546781491754)\n",
      "===============================================\n",
      "UCB1\n",
      "Task 1 10.91\n",
      "Task 2 17.05\n",
      "Task 3 44.42\n",
      "Task 4 31.62\n",
      "('Total reward ', 2.7891283524904202)\n",
      "===============================================\n"
     ]
    }
   ],
   "source": [
    "run(100, False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}